INTERNATIONAL ORGANISATION FOR STANDARDISATIONORGANISATION INTERNATIONALE DE NORMALISATION

ISO/IEC JTC 1/SC 29/WG 11CODING OF MOVING PICTURES AND AUDIO

ISO/IEC JTC 1/SC 29/WG 11 N7267Poznań, PL – July 2005

Source: Leonardo Chiariglione Title: Report of 73rd meetingStatus:

Report of 73rd meeting

1 Opening The 73rd WG11 meeting was held in Poznań at the kind invitation of the Polish National Body and hosted by Poznań University of Technology, Institute of Electronics and Telecommunications, Division of Multimedia Telecommunications and Radioelectronics.

2 Roll call of participants Annex 1 gives the list of participants

3 Approval of agendaAnnex 2 gives the approved agenda

4 Allocation of contributionsAnnex 3 gives the list of submitted contributions

5 Communications from Convenor There were no specific communications

6 Report of previous meeting This was approved

7 Processing of NB Position PapersThese were considered and responses provided where appropriate

8 Work plan

8.1 Media coding

8.1.1 New levels for Simple Profile Conformance The following documents were approved

1

7306 Disposition of Comments on ISO/IEC 14496-4:2004 FPDAM97307 Text of ISO/IEC 14496-4:2004 FDAM9 (AVC Frext)

8.1.2 Scalable Video Coding The following documents were approved

7310 Working Draft 3 of ISO/IEC 14496-10:200x/AMD1 Scalable Video Coding

7311 Joint Scalable Video Model (JSVM) 37312 JSVM 3 Software

8.1.3 Audio Lossless Coding The following documents were approved

7363 DoC of 14496-3:2001/FPDAM 4, Audio Lossless Coding (ALS), new audio profiles and BSAC extensions

7364 Text of 14496-3:2001/FDAM 4, Audio Lossless Coding (ALS), new audio profiles and BSAC extensions

7371 Draft Performance and Complexity of MPEG Lossless Audio Coding Standards7372 Workplan for Lossless Coding

8.1.4 Audio Scalable Lossless Coding The following documents were approved

7365 DoC of 14496-3:2001/FPDAM 5, Scalable Lossless Coding (SLS)7366 Text of 14496-3:2001/FDAM 5, Scalable Lossless Coding (SLS)7371 Draft Performance and Complexity of MPEG Lossless Audio Coding Standards7372 Workplan for Lossless Coding

8.1.5 Low Delay AAC ProfileThe following documents were approved

7367 Request for Amendment 14496-3:2005/AMD 17368 Text of 14496-3:2005/PDAM 1, Low Delay AAC Profile

8.1.6 Parametric Coding for High Quality Audio ConformanceThe following documents were approved

7460 Request for ISO/IEC 14496-4:2004 AMD13 Parametric Coding for High Quality Audio Conformance

7461 Text of ISO/IEC 14496-4:2004 PDAM13 Parametric Coding for High Quality Audio Conformance

8.1.7 MPEG SurroundThe following documents were approved

7387 WD 2 for MPEG Surround

7388 Performance of MPEG Surround RM Technology

7389 Workplan for MPEG Surround

2

7391 Study on Profiles for MPEG Surround

8.1.8 AFX Extensions The following documents were approved

7399 DoC on ISO/IEC 14496-16/FPDAM17400 Text of ISO/IEC 14496-16/FDAM1 (Morphing & Textures)7401 WD 2.0 of 14496-16 2nd edition7402 WD 2.0 of 14496-16/AMD2 (Geometry & Shadow) 7403 Text of ISO/IEC 14496-16:2004/COR2 (extrinsic camera parameters)7404 Text of ISO/IEC 14496-16:2004/DCOR3 (mesh grid update)7405 AFX Core Experiments Description

8.1.9 Streaming Text Format The following documents were approved

7478 DoC on ISO/IEC 14496-17/FCD7479 Text of ISO/IEC 14496-17/FDIS

8.1.10 Generic inverse DCT specificationThe following documents, caused by the said specification, were approved 7289 Text of ISO/IEC 11172-2 DCOR47291 Text of ISO/IEC 11172-4 DCOR47293 Text of ISO/IEC 13818-2 DCOR47295 Text of ISO/IEC 13818-4 DCOR47292 Study of ISO/IEC 11172-6 FCD

8.1.11 Video Coding Tool Repository The following documents were approved

7329 Study of Video Coding Tools Repository V5.07330 VCTR Textual Description V4.07331 VCTR Software V3.0

8.1.12 3D AV Coding The following documents were approved

7282 Requirements on Multiview Video Coding 7327 Call for Proposals on Multi-View Video Coding

8.1.13 Open Font Format The following documents were approved

7484 Request for a new part of ISO/IEC 144967485 Text of ISO/IEC 14496-22/CD

8.1.14 Exploration on wavelet video coding The following documents were approved

3

7333 Description of Exploration Experiments in Wavelet Video Coding7334 Wavelet Codec Reference Document and Software Manual V1.0

8.1.15 Scalable audio and speech coding The following document was approved

7392 Brief Report on Call for Information on Scalable Speech and Audio Coding

8.1.16 Fixed point implementation of DCT/IDCT The following document was approved

7335 Call for Proposals on Fixed-point 8x8 IDCT and DCT

8.1.17 Colour spaces The following document was approved

7283 Report of Activities on Supporting New Colour Spaces in MPEG Standards

8.2 Composition coding

8.2.1 Lightweight Scene Representation The following documents were approved

7480 Study Text of ISO/IEC 14496-20/FCD7481 TuC for ISO/IEC 14496-207482 Validation Experiment on LASeR Binary Encoding7483 Validation Experiment on LASeR streamSource

8.2.2 Symbolic Music Representation The following documents were approved

7375 Core Experiment Methodology for Symbolic Music Representation7376 Workplan for SMR Core Experiment on Traditional Korean Music7377 Workplan for SMR Core Experiment on Braille and Spoken Music7378 Text of Working Draft for Symbolic Music Representation (SMR)

8.3 Description coding

8.3.1 Visual Descriptor Extensions The following documents were approved

7316 Study Text of ISO/IEC 15938-3:2002/FPDAM2 7317 MPEG-7 Visual XM Document version 25.07318 Description of Core Experiments for MPEG-7 New Visual

Extensions

8.3.2 Audio Descriptor Extensions The following documents were approved

7384 DoC on 15938-4:2002/PDAM 4 (Audio High-Level Description Extensions)

4

7385 Text of 15938-4:2002/FPDAM 4 (Audio High-Level Description Extensions)7386 Workplan for MPEG-7 Audio

8.3.3 Extraction and use of descriptors and description schemes The following documents were approved

7321 Text of ISO/IEC TR 15938-8:2002 COR17322 Text of ISO/IEC TR 15938-8:2002

PDAM2

8.4 Systems support

8.4.1 3D Compression Profile DescriptorsThe following documents were approved

7517 Request for minor enhancement ISO/IEC 14496-1:20047472 Text of ISO/IEC 14496-1:2004/PDAM2 3D Compression Profile Descriptor

8.5 IPMP

8.5.1 DII Related IdentifiersThe following documents were approved

7424 DoC of PDAM/1 on DII Related Identifiers7425 Study of PDAM/1 on DII Related Identifiers

8.5.2 REL Profiles The following documents were approved

7428 Request for Amendment 1 of ISO/IEC 21000-57429 PDAM/1 MPEG-21 REL Profiles

8.5.3 Rights Data DictionaryThe following documents were approved

7430 Request for Amendment 1 of ISO/IEC 21000-67431 PDAM/1 ISO/IEC21000-6 Digital Item Identifier Relationship Types7432 Workplan for CE on RDD Abstraction and Resource

8.6 Digital Item

8.6.1 DIA Conversions and Permission Extensions The following documents were approved

7434 Study of ISO/IEC 21000-7 FPDAM/17435 TuC for 21000-7 MPEG-21 DIA Amendment(s) v3.07436 WD 1.0 of ISO/IEC 21000-7 AMD/27437 DIA Reference Software Status and Workplan 7438 Workplan for CE on Adaptation QoS and BSDLink extensions 7439 Workplan for CE on BSDL Extensions

5

8.6.2 Event Reporting The following documents were approved

7443 DoC on ISO/IEC 21000-15 CD MPEG-21 ER7444 FCD of ISO/IEC 21000-15 MPEG-21 ER

8.6.3 Fragment Identification of MPEG Resources The following documents were approved

7445 DoC on ISO/IEC 21000-17 CD MPEG-21 Fragment Identification of MPEG Resources7446 ISO/IEC 21000-17 FCD MPEG-21 Fragment Identification of MPEG Resources

8.7 Transport and File Format

8.7.1 Transport of Streaming Text over MPEG-2The following documents were approved

7470 Request of Amendment of ISO/IEC 13818-1:20007471 Text of ISO/IEC 13818-1:2000/PDAM 6

8.7.2 Digital Item Streaming The following documents were approved

7278 Requirements, Terminology and Use Cases for DI Streaming 7447 Working Draft of MPEG-21 DI Streaming7448 Workplan for CE on Fragmentation for DI Streaming7449 Workplan for CE on DI Mapping to Delivery channels

8.8 Multimedia architecture

8.8.1 MPEG-J GFX

The following documents were approved

7407 DoC on ISO/IEC 14496-21/CD7408 Text of ISO/IEC 14496-21/FCD

8.8.2 MPEG Multimedia Middleware

The following documents were approved

7494 WD2.0 of ISO/IEC 23004-1 Architecture7495 WD2.0 of ISO/IEC 23004-2 Multimedia API7496 WD2.0 of ISO/IEC 23004-3 Component Model7497 WD2.0 of ISO/IEC 23004-4 Resource and Quality Management7498 WD2.0 of ISO/IEC 23004-5 Component Download7499 WD2.0 of ISO/IEC 23004-6 Fault Management7500 WD2.0 of ISO/IEC 23004-7 System Integrity Management

8.8.3 Digital Item Processing

The following documents were approved

6

7441 Request for Amendment 1 of ISO/IEC 21000-107442 PDAM/1 of ISO/IEC21000-10 DIP: Additional C++ Bindings

8.9 Reference implementation

8.9.1 AVC Fidelity Range Extensions Reference Software The following documents were approved

7308 Disposition of Comments on ISO/IEC 14496-5:2001 FPDAM87309 Text of ISO/IEC 14496-5:2001 FDAM8 (AVC Frext)

8.9.2 Parametric Coding for High Quality Audio Reference Software The following documents were approved

7462 Request for ISO/IEC 14496-5:200x AMD10 Parametric Coding for High Quality Audio Reference Software

7463 Text of ISO/IEC 14496-5:200x PDAM10 SSC, DST, ALS, SLS Reference Software

8.9.3 Reference Hardware Description The following documents were approved

N7273 Status of HDL submissions and commitments for MPEG-4 Part-9N7275 Study of Part ISO/IEC PDTR 14496-9:2005 2nd Edition Reference Hardware Description

8.9.4 MPEG-7 Systems Reference SoftwareThe following document was approved

7488 MPEG-7 Systems reference software workplan

8.10 Application formats

8.10.1 Application Format Framework The following document was approved

7281 PDTR for ISO/IEC 23000-1

8.10.2 PhotoPlayer Application Format The following documents were approved

7323 Request for Subdivision of ISO/IEC 23000: Part 3 Photo Player Multimedia Application Format

7324 WD 2.0 of 23000-3 Photo Player Multimedia Application Format

9 Liaison matters The following liaison statements were issued

7351 Liaison Statement to ITU-R SG 6/WP 6A on ITU-R BS.1387-17352 Liaison Statement to ITU-R SG 6 WP 6A on objective evaluation of spatial audio

7

reproduction7353 Liaison Statement to SMPTE on OpenType® and ISO/IEC 14496-227354 Liaison Statement to ECMA on Universal 3D File Format7355 Liaison Statement to JTC1 SC 377356 Liaison Statement to 3GPP7357 Liaison Statement to IEC TC1007358 Liaison Statement to TVAF7359 Liaison Statement to OMA7360 Liaison Statement to DMP7361 Liaison Statement to TC 42 WG187362 Liaison Statement to Khronos OpenMax7413 Liaison Statement to DVB TM-AVC7414 Liaison Statement to ATSC S67415 Liaison Statement to SCTE DVS7416 Liaison Statement to DVD Forum WG-17417 Liaison Statement to Blu-ray Disc Association TEG7459 Liaison to 3GPP on loss resilience video experiments

10 Organisation of this meeting

10.1 Tasks for subgroups The following tasks were assigned

Subgroup S P A TasksRequirements

4 16 3D compression profiles4 20 Laser profiles7 JPSearch21 5 REL Profiling

6 RDD Implementation Issues18 DI streaming

A 1 PDTR2 Encrypted Audio player MAF3 Photo player MAF

MAF landscapeBX Content-Based Search Framework

Systems2 1 3rd edition4 15 1 AVC File Format extensions for FRExt

17 Streaming Text Format20 Lightweight Scene Representation22 Font Format Representation

7 1 2 XML Structure Binarization and StreamingA MAF landscapeB 1 XML Structure Binarization and StreamingX MPEG Multimedia Middleware

MDS21 3 Related identifiers

8

4 IPMP Components6 RDD Implementation Issues7 DIA Conversions and Permission Extensions

10 1 Digital Item Processing15 Event Reporting17 Fragment Identification of MPEG Resources18 Digital Item Streaming

MPEG-21 WorkshopA 2 Encrypted Audio player MAF

3 Photo player MAFMAF landscape

Video1 6 Video IDCT Specification4 2 Removal of technology unsupported by conformance

4 FRExt conformance5 FRExt reference software

10 2 Scalable Video Coding7 3 2 New Visual Extensions

4 8 Extraction and use of Ds and DSsAB 2 Fixed-point 8x8 IDCT and DCT CfPX Colour spaces

3DAV Coding CfPVideo Coding Tools RepositoryWavelet Video CodingFuture Directions in Video Compression

Audio4 Lossless Audio Coding7 New Audio ExtensionsA 2 Encrypted Audio player MAFX Spatial Audio Coding

Symbolic Music RepresentationScalable Audio and Speech Coding

SNHC4 4 12 Conformance of AFX Amd 1

5 9 Reference software of AFX Amd 116 1 AFX21 MPEG-J GFX

XIntegration

1 4 Conformance5 Reference Software

2 4 Conformance5 Reference Software

4 4 Conformance5 Reference Software

7 6 Reference Software7 Conformance

21 8 Reference Software

9

Test Bed for MPEG-21 Resource Delivery14 Conformance

A 2 Reference Software2 Conformance

Test4 10 2 SVC sequencesX 3DAV Coding CfP

ISG4 Reference Hardware Description

Liaison124721ABX

10.2 Joint meetings The following joint meetings were held

Groups What Where Day TimeReq, mds, sys DI streaming Req Mon 16:30-18:00Req, mds REL profiling Req Tue 09:00-11:00Req, vid Photo player Vid Tue 11:00-11:30Req, sys Laser, font format Req Tue 12:00-13:00Req, snh 3D compression profiling snh Tue 14:00-15:00Req, mds RDD implementation Req Tue 15:00-16:00Req, aud, mds, sys Protected mMusic player maf Req Tue 16:00-18:00Req, vid Colour spaces/sampling, multiview Req Wed 12:00-13:00Vid, jvt, sys SVC systems issues Vid Wed 14:00-16:00

MPEG-21 tutorial REq Wed 14:00-17:45Req, Vid Photo player, CB Search Vid-7 Wed 14:00-15:00Req, Aud MPEG-4 Audio profiling Aud Wed 17:00-17:30Vid, isg IDCT Vid Wed 17:00-17:45Req, Mds, sys MPEG-21 Systems Req Thu 09:00-10:00Req, Mds, sys DI streaming Req Thu 10:00-12:00Req, snh 3D compression profiling Snh Thu 12:30-13:00Req, mds REL profiling Req Thu 14:00-15:00Snh, int MPEG-4 ref sw Snh Thu 14:00-15:00Req, mds, aud, sys Protected MP MAF Req Thu 15:00-16:00

11 Administrative matters

11.1 Schedule of future MPEG meetings The following meeting plan was adopted

10

Meeting City Country yy mm dd73rd Poznań PL 05 07 25-2974th Nice FR 05 10 17-2175th Bangkok TH 06 01 16-2076th Genève CH 06 04 03-0777th Klagenfurt AT 06 07 17-2178th Hangzhou CN 06 10 23-2779th Marrakech MA 07 01 15-19

WG11 members are advised that the April 2006 meeting awaits final confirmation. In case such a meeting cannot be held as planned fall-back solutions will be discussed and proposed by the HoDs.

11.2 Promotional activities The following 2-page descriptions of MPEG technologies were approved for posting on the MPEG web site

7290 Introduction to MPEG-1 Video 7294 Introduction to MPEG-2 Video 7297 Introduction to MPEG-4 Part 2 Video (rectangular)7298 Introduction to MPEG-4 Part 2 Video (arbitrary shape)7314 Introduction to MPEG-4 Part 10 Advanced Video Coding 7315 Introduction to SVC Extension of Advanced Video Coding 7319 Introduction to Low-level Visual Description Tools7320 Introduction to High-level Visual Description Tools7325 Introduction to Photo Player Multimedia Application Format7326 Introduction to Accuracy Requirements for 8x8 IDCT7328 Introduction to Multi-View Video Coding7332 Introduction to Video Coding Tools Repository7336 Introduction to Fixed-point Implementation of DCT/IDCT7379 Symbolic Music Representation in MPEG, an overview7390 Tutorial on MPEG Surround7409 MPEG-J GFX white paper7410 MPEG-J white paper7418 Introduction to MPEG-7 MDS – an overview 7420 Introducing ISO/IEC 21000 Schema Files7422 Introducing MPEG-21 Digital Item Declaration7423 Introducing ISO/IEC 21000-3 DII7427 Introducing the MPEG REL 7433 Introducing the MPEG-21 RDD 7440 Introducing the MPEG-21 Reference Software7455 Overview of Evaluation Methods for Persistent Association Technologies7456 Overview of FBA (Face and Body Animation)7457 Overview of 3DMC (3D Mesh Coding)7458 Overview of AFX (Animation Framework eXtension)7464 Description of Spectral and Replication7465 Description of Lossless Coding of Over Sample Audio7466 Description of MPEG-7 Low level descriptors7467 Description of MPEG-7 high level descriptors 7468 Description of Parametric Coding of High Quality Audio

11

7502 White Paper on MPEG-1 and MPEG-2 Systems7503 White Paper on MPEG-2 IPMP7504 White Paper on MPEG-4 Systems7505 White Paper on MPEG-4 IPMP7506 White Paper on MPEG-4 BIFS7507 White Paper on LASeR7508 White Paper on Font Compression and Streaming7509 White Paper on MPEG-7 Systems7510 White Paper on Multimedia Middleware7515 White Paper on Streaming Text7519 White Paper on Open Font Format

12 Planning of future activitiesThe following ad hoc groups were established

7411 Ad Hoc Group on AFX documents, CEs, and software7513 Ad Hoc Group on MPEG File Formats7514 Ad Hoc Group on MPEG Multimedia Middleware7512 Ad Hoc Group on Scene Representation7511 Ad Hoc Group on XML Structure Binarization and Streaming7405 AFX Core Experiments Description7342 AHG on 3D Video Coding 7393 AHG on Audio Standards Maintenance7288 AHG on Colour Spaces7286 AHG on Content Based Search Framework7341 AHG on Description Tools for MPEG-7 New Visual Extensions7344 AHG on Exploration in Wavelet Video Coding 7398 AHG On Exploration of Scalable Audio and Speech Coding7285 AHG on MAFs7339 AHG on Maintenance of MPEG-4 Visual related Documents, Reference Software and

Conformance7340 AHG on Maintenance of MPEG-7 Visual related Documents and Reference Software7284 AHG on MPEG Dictionary7395 AHG on MPEG Surround7451 AHG on MPEG-21 DIA7450 AHG on MPEG-21 DIP7454 AHG on MPEG-21 DIStreaming7452 AHG on MPEG-21 Event Reporting7453 AHG on MPEG-21 IPMP Components7396 AHG on MPEG-4 Lossless Audio Coding7274 AHG on MPEG-4 Part 9: Reference Hardware Description Phase 1 and 2.7394 AHG on MPEG-7 Audio7412 AHG on MPEG-J Graphics Framework eXtensions7345 AHG on Organization of Workshop on Future Directions in Video Compression7346 AHG on Photo Player MAF7397 AHG on Symbolic Music Representation7343 AHG on Video Coding Tools Repository 7338 AHG on Video IDCT Specification 7287 AHG on XML Fragment Requests

12

13 Resolutions of this meeting These were approved (N7265)

14 A.O.B There was no other business

15 ClosingThe meeting closed on 2005/07/29T22:15 with thanks to the hosting organisation

13

Annex 1Attendance list

First name Last name Affiliation Cou.Michael Ransburg Klagenfurt University ATChristian Timmerer Klagenfurt University ATMyriam Amielh CiSRA AUIan Burnett University of Wollongong AUGerrard Drury University of Wollongong AUJoseph Thomas-Kerr University of Wollongong AUFrederik De Keukelaere Ghent University BEWesley De Neve Ghent University BESaar De Zutter Ghent University BEIan Bormans IMEC BESalomie Alexandru Vrije Universiteit Brussel BEJan Lievens Vrije Universiteit Brussel BEPeter Schelkens Vrije Universiteit Brussel BERichard Cam PMC-SIERRA CAIhab Amer University of Calgary CATouradj Ebrahimi EPFL CHGiorgio Zoia EPFL - ITS CHThomas Di Giacomo MIRALab, University of Geneva CHJulien Reichel VisioWave S.A. CHDiego Santa Cruz VisioWave S.A. CHFrancesco Ziliani VisioWave S.A. CHLianhuan Xiong Huawei Tech. Corp., Ltd. CNWen Gao Institute of Computing Technology, CAS CNJizheng Xu Microsoft Research Asia CNChen Ying Thomson Beijing CR CNLihua Zhu Thomson R&D (Beijing) CNLu Yu Zhejiang University CNMartin Dietz Coding Technologies GmbH DEAndreas Schneider Coding Technologies GmbH DEBernhard Feiten Deutsche Telekom T-Systems ENPS DEMatthias Gruhne Fhg IDMT DEDetlev Marpe Fraunhofer HHI DEKarsten Müller Fraunhofer HHI DEThomas Schierl Fraunhofer HHI DEHeiko Schwarz Fraunhofer HHI DEAljoscha Smolic Fraunhofer HHI DEThomas Wiegand Fraunhofer HHI DEHarald Fuchs Fraunhofer IIS DERalf Geiger Fraunhofer IIS DEBernhard Grill Fraunhofer IIS DEJuergen Herre Fraunhofer IIS DEJohannes Hilpert Fraunhofer IIS DEStefan Kraegeloh Fraunhofer IIS DENikolaus Rettelbach Fraunhofer IIS DERalph Sperschneider Fraunhofer IIS DEHerbert Thoma Fraunhofer IIS DEThomas Wedi Panasonic DE

14

Steffen Wittmann Panasonic European Laboratories DEUlrich Benzler Robert Bosch GmbH DEMarco Boltz Robert Bosch GmbH DEJens-Rainer Ohm RWTH Aachen University DEMarkus Beermann RWTH Aachen University DEMathias Wien RWTH Aachen University DEGero Bäse Siemens AG DEAndreas Hutter Siemens AG DEAmon Peter Siemens AG DETilman Liebchen Technical University of Berlin DEThomas Rathgen Technische Universität Ilmenau DETobias Oelbaum Technische Universität München DEJohannes Boehm THOMSON DEIngo Wolf T-Systems International GmbH DEJoern Ostermann Universität Hannover DEThomas Kummer -Hardt Aid International ESGabriel Fernandez Ramon Llull University - La Salle, Digital TV

CentES

Marc Gauvin sDAe ESMarcos Avilés

RodrigálvarezUniversidad Politécnica de Madrid ES

Francisco Morán Burgos Universidad Politécnica de Madrid ESJaime Delgado Universitat Pompeu Fabra ESEva Rodriguez Universitat Pompeu Fabra (UPF) ESYe-Kui Wang Nokia Research Center FIFabrice Le Leannec Canon Research Centre France SAS FRSebastian Brangoulo ENST FRCyril Concolato ENST FRBenoit Pellan ENST FRPhilippe De Cuetos Expway FRClaude Seyrat Expway FRAbdellatif Benjelloun France Telecom division R&D FRSylvain DEVILLERS France Telecom division R&D FRPatrick Gioia France Telecom division R&D FRJean-Bernard Rault France Telecom division R&D FRStephane Pateux France Telecom division R&D FRNathalie Cammas France Telecom division R&D FRDavid Virette France Telecom division R&D FRIsabelle Amonou France Telecom division R&D FRPierieck Philipe France Telecom division R&D FRAlexandre Cotarmanac’h France Telecom division R&D FRMarius Preda INT FRGregoire Carpetier IRCAM FRArnaud Bourge Philips FRMarc GUEZ VUCHER SCPP FRJean-Claude Dufourd Streamezzo FRAVARO Olivier STREAMEZZO FRJerome Pieron Thomson FRFrancois Eduard Thomson FRZvi Lifshitz Optibase Ltd. ILYotam Toib Quix Technologies IL

15

Ran Boneh Quix Technologies ILKalyan Kaipa Samsung India Software Operations INLeonardo Chiariglione CEDEO Sas ITMaurizio Campanai Exitech ITStefano Tubaro Politecnido di Milano ITGiovanni Cordara Telecom Italia LAB ITRiccardo Leonardi University of Brescia ITPaolo Nesi University of Florence DSi-DISIT ITPierfrancesco Bellini University of Florence DSi-DISIT ITAndrea Valotti University of Florence DSi-DISIT ITTakashi Itoh Fujitsu Laboratories Ltd. JPMark Callow HI Corporation JPTomokazu Murakami Hitachi JPYukiko Ogura IPSJ/ITSCJ JPYoshinori Sugihara JEITA JPYoshiaki Takagi Matsushita Electric (Panasoinc) JPKenji Takita Matsushita Electric (Panasonic) JPTAKAFUMI UENO Matsushita Electric (Panasonic) JPTAKESHI NORIMATSU Matsushita Electric Industrial Co., Ltd. JPKohtaro Asai Mitsubishi Electric Corporation JPTokumichi Murakami Mitsubishi Electric Corporation JPHirofumi Nishikawa Mitsubishi Electric Corporation JPYoshihisa Yamada Mitsubishi Electric Corporation JPToshiaki Fujii Nagoya University JPMasayuki Tanimoto Nagoya University JPYoshihiro Miyamoto NEC JPToshiyuki Nomura NEC JPAkio Yamada NEC JPTakahiro Kimoto NEC Corporation JPOsamu Shimada NEC Corporation JPYukihiro Bandoh NTT JPNoboru Harada NTT JPMasaki Kitahara NTT JPTakehiro Moriya NTT JPYoshiyuki Yashima NTT JPHideaki Kimata NTT Advanced Technology JPTerumasa Aoki Research Center for Advanced Science and

TechnologyJP

Tomoyuki Yamamoto Sharp Corporation JPMark Veltman Sony JPTeruhiko Suzuki Sony Corp. JPMasato Shima Texas Instruments Japan Limited JPTakuyo Kogure The University of Tokyo JPTakanori Senoh The University of Tokyo JPTakeshi Chujoh Toshiba Corporation JPSatoshi Ito TOSHIBA CORPORATION JPMayumi Koike U-Tokyo JPHiroya Nakamura Victor Company of Japan, Limited JPSatoru Sakazume Victor Company of Japan, Limited JPHIROSHI WATANABE Waseda University JPEuee S Jang College of Information and Communication KR

16

Sunyoung Lee College of Information and Communication KRWoo-Shik Kim Computing Lab. Samsung AIT KRTaehyun Kim DRM Inside KRHogab Kang DRM Inside KRJinwoo Hong Electrinics and Telecommunications Research

InstituteKR

Byung-Gyu Kim Electronics and Telecommunications Research Institute

KR

Seon-Tae Kim Electronics and Telecommunications Research Institute

KR

Kyeongok Kang Electronics Telecomminications Reserach Institute

KR

Eun-Young Chang ETRI KRYongju Cho ETRI KRHyon-Gon Choo ETRI KRYesun Joung ETRI KRDaehee Kim ETRI KRKyuheon Kim ETRI KRJae-Gon Kim ETRI KRSeong Won Ryu ETRI KRJeongil Seo ETRI KRYoungho Suh ETRI KRWeon Geun Oh ETRI KRSooJun Park ETRI KRHaechul Choi ETRI KRSeyoon Jeong ETRI KRJung Won Kang ETRI KRGun Bang ETRI KRJong Mo Sung ETRI (Electronics and Telecommunication

Research IKR

Yo-Sung Ho GIST KRSeung-Uk Yoon Gwangju Institute of Science and Technology KRHeung Soo Park HANWHA S&C CO., LTD KRBeack Seung Kwon ICU KRHahn Min Soo ICU (Information and Communication University KRTae Meon Bae Information and Communication University KRTruong Cong Thang Information and Communication University KRHendry Hendry Information and Communications University KRMunchurl Kim Information and Communications University KRSeungji Yang Information and Communications University KRJeongyeon Lim Information and Communications University

(ICU)KR

Dong-Seok Leong Inha University KRIlHong Shin KAIST KRYONGSOO Choi Kangwon National University KRJae Yeal Nam Keimyung University KRYae Yeal Nam Keinyang University KRHyok Song KETI KRByeongho Choi KETI KRYong-Hwan Kim KETII KRJe Woo Kim Ketz KR

17

Doug Young Suh KHU KRNam Hyeong Min Korea University KRKim Nam Hyeong Korea university KRKeunSoo Park Koreal Broadcasting System KRSun-Jeong Kim KT KRJongheum Kim KT KRJung-Sup Ryu KT KRDong gy Sim Kwangwoon University KRMin Woo Park Kyung Hee University KRKyung Il Lee Kyung Hee University KRGwang Hoon Park Kyung Hee University KRHyen-O Oh LG Electronics KRHee-Suk Pang LG Electronics KRSeung-Wook Park LG Electronics KRJeongHyu Yang LG Electronics KRSun-Mi Kim LG Electronics KRKyeong-Su Yim LG Electronics KRByeong-Moon Jeon LG Electronics Inc. KRJi Ho Park LG Electronics Inc. KRDoe Hyun Yoon LG Electronics Inc. KRTae Hyeon Kim LG Electronics Institute of Technology KRDong Soo Kim LG Electronics Ltd. KRJae hyun Lim LG Electronics. KRYoung-Kwon Lim net&tv Inc. KRJong-Lak Kim Nextreaming Corporation KRWoo Cheol Sung PIXTREE KRSunghee Park Pixtree technologies KRSang Wook Kim Samsung KRHyun Mun Kim Samsung KRJeong-Hwan Ahn SAMSUNG A.I.T. KRSang-Kyun Kim Samsung Advanced Institute of Technology KRKyungseok Lee Samsung Advanced Institute of

Technology(SAIT)KR

Mahnjin Han Samsung AIT KRJungHoe Kim Samsung AIT KRWoo-Jin Han Samsung Electronics KRBae-Keun Lee Samsung Electronics KRKue-Hwan Sihn Samsung Electronics KRHa Tae Hyeun Samsung Electronics KRSong Young Joo Samsung Electronics KRYun-ju Yu Samsung Electronics Co. Ltd KREunmi Oh Samsung Electronics Co. Ltd KRKyohyuk Lee Samsung Electronics Co., Ltd. KRShin Sungchol Samsung Electronics. Co,Ltd KRSangchang Cha Samsunglectronics, co. Ltd, KRChan Yul Park School Of Electrical Engineering and Computer

ScieKR

Han Sae Song School Of Electrical Engineering and Computer Scie

KR

Hae Kwang Kim Sejang University KRJae Ho Hur Sejong Univ. DMS Lab. KR

18

Yung-Lyul Lee Sejong University KRYung Gi Lee Sejong University KRChang-Yul Kim Sejong University KRKi-Hun Han Sejong University KRJong-Ki Han Sejong University KRJie JIA Sejong University KRDae-Yeon Kim Sejong University KRIn Yung Choi Seoul national University KRWOONG IL Choi Sungkyunkwan University KRByeungwoo Jeon Sungkyunkwan University KRKyunghee Ji SUV KRYong-tae Kim Yonsei University KRSangyoun Lee Yonsei University KRKwang Hoon Sohn Yonsei University KRNeil McKenzie FNB NLWerner Oomen Philips Applied Technologies NLJan van der Meer Philips Electronics NLJean H.A. Gelissen Philips Research NLHugh Maaskant Philips Research NLAndrew Tokmakoff Telematica Instituut NLPeder Drege Adactus AS NORunar J. Kolberg Adactus AS NOWładysław Skarbek Politechnika Warszawska, Instytut

RadioelektronikiPL

Karol Wnukowicz Politechnika Warszawska, Instytut Radioelektroniki

PL

Ewa Łukasik Poznań University of Technology PLMarek Domański Poznań University of Technology PLAdam Łuczak Poznań University of Technology PLMaciej Bartkowiak Poznań University of Technology PLPaweł Piotrowski Telekomunikacja Polska SA PLLuis A. da Silva Cruz Dep. Electrical and Computer Engineering PTPedro Carvalho INESC Porto PTFernando Pereira Instituto Superior Técnico PTKristofer Kjörling Coding Technologies AB SEHeiko Purnhagen Coding Technologies AB SEPer Fröjdh Ericsson SEAnisse Taleb Ericsson AB SETe Li Institute for Infocomm Research SGKeng Pang Lim Institute for Infocomm Research SGXiao Lin Institute for Infocomm Research SGZhongkang Lu Institute for Infocomm Research SGSusanto Rahardja Institute for Infocomm Research SGRongshan Yu Institute for Infocomm Research SGTK Tan NTT DoCoMo, Inc SGKok Seng Chong Panasonic Singapore Laboratories SGChong Soon Lim Panasonic Singapore Laboratories SGMen Huang Lee Panasonic Singapore Laboratories (PSL) SGZhongyang Huang Panasonic Singapore Laboratories Pte Ltd SGSua Hong Neo Panasonic Singapore Laboratories Pte Ltd SGChinPhek Ong Panasonic Singapore Laboratories Pte Ltd SG

19

ChakJoo Lee Panasonic Singapore Laboratory SGMike Nilsson BT UKMartin Russ BT Group plc UKRobert O'Callaghan Mitsubishi Electric ITE VI-Lab UKMiroslaw Bober Mitsubishi Electric ITE-VIL UKLeszek Cieplinski Mitsubishi Electric ITE-VIL UKRay Taylor NDS UKCatherine Grant Nine Tiles UKTanya Beech QinetiQ UKChris Barlas Rightscom Limited UKNiels Rump Rightscom Limited UKSimon Watt Universal Music International UKScott Foshee Adobe Systems Inc USDavid Singer Apple USSchuyler Quackenbush Audio Research Labs USWade Wan Broadcom Corporation USMukta L Kar Cable television Labs USKevin Short Chaoticom Technologies/Groove Mobile USXin Wang ContentGuard, Inc. USMichael Horowitz CoVi Technologies USMatt Fellers Dolby Laboratories USWalt Husak Dolby Labs USPankaj Topiwala FastVDO USYolanda Prieto Freescale Semiconducrior Inc. USPeter P. Schirling IBM Research USYi-Jen Chiu Intel Corp. USReha CIVANLAR Layered Media USAlexandros Eleftheriadis Layered Media USOfer Shapiro Layered Media USCaspar Horne Metta Technology USFlorian Pestoni Microsoft USAndrew Tescher Microsoft USMike Ksar (1550) Microsoft Co. USGary Sullivan Microsoft Corp USMikael Bourges-Sevenier Mindego Inc. USAnthony Vetro Mitsubishi Electric USEmin Martinian Mitsubishi Electric Research Labs USHuifang Sun Mitsubishi Electric Research Labs USJose Roberto Alvarez Mobilygen Corporation USVladimir Levantovsky Monotype Imging Inc. USAjay Luthra Motorola USSam Narasimhan Motorola USMarina Bosi MPEG LA LLC USYi-Shin Tung National Taiwan University USChun-Jen Tsai NCTU/ITRI USJeroen Bekaert NISO - LANL USMike Rubinfeld NIST National INST of Standards and

TechnologyUS

Wo Chang NIST National INST of Standards and Technology

US

Vidia Seltur Nokia US

20

Justin Ridge Nokia USXianglin (Shawn) Wang Nokia USYiliang Bao Nokia Inc. USMarta Karczewicz Nokia Inc. USHsin-Hao Chen OES/ITRI USViji Raveendran QUALCOMM Incorporated USYuriy Reznik QUALCOMM Incorporated USPhoom Sagetong Qualcomm Incorporated USJun Zhang Santa Clara University USMichael Isnardi Sarnoff Corporation USArturo Rodriguez Scientific-Atlanta Inc. USAndrew Segall Sharp Labs USShijun Sun Sharp Labs of America USMohammed Zubair

Visharam Sony Electronics US

Viswanathan Swaminathan Sun Microsystems Laboratories USGregory W. Cook Thomson USCristina Gomila Thomson USHaoping Yu Thomson USJill Boyce Thomson US

Annex 2Agenda

Agenda item 1. Opening 2. Roll call of participants 3. Approval of agenda 4. Allocation of contributions 5. Communications from Convenor 6. Report of previous meeting 7. Processing of NB Position Papers8. Work plan 8.1 Media coding 1.1. New levels for Simple Profile 1.2. New levels for Simple Profile Conformance 1.3. AVC Fidelity Range Extensions Conformance 1.4. Scalable Video Coding 1.5. Audio Lossless Coding 1.6. Audio Scalable Lossless Coding 1.7. Spatial Audio Coding 1.8. AFX Extensions 1.9. Streaming Text Format 1.10. Generic inverse DCT specification 1.11.. Video Coding Tool Repository 1.12. 3D AV Coding 1.13. Open Font Format 1.14. Exploration on wavelet video coding 1.15. Scalable audio and speech coding 1.16. Fixed point implementation of DCT/IDCT

21

1.17. Colour spaces 8.2 Composition coding 2.1. Lightweight Scene Representation 2.2. Symbolic Music Representation 8.3 Description coding 3.1. Visual Descriptor Extensions 3.2. Audio Descriptor Extensions 3.3. Extraction and use of descriptors and description schemes 8.4 Systems support 4.1 Text Profile Descriptors 4.2 MPEG-7 Random Access in BiM 4.2 XML Binarisation 8.5 IPMP 5.1. MPEG-21 IPMP Components 5.2. REL Profiles 5.3. Rights Data Dictionary 8.6 Digital Item 6.1. DIA Conversions and Permission Extensions 6.2. Event Reporting 6.3. Fragment Identification of MPEG Resources 6.4. MPEG-21 Conformance 6.5. Digital Item Identification 8.7 Transport and File Format 7.1. AVC File Format extensions for FRExt 7.2. SVC File Format extensions 7.3. Digital Item Streaming 8.8 Multimedia architecture 8.1. MPEG-J GFX 8.2. MPEG Multimedia Middleware 8.3. Digital Item Processing 8.4. Search Framework 8.9 Reference implementation 9.1. AVC Fidelity Range Extensions Reference Software 9.2. Reference Hardware Description – phase I MPEG-4 Visual 9.3. Reference Hardware Description – phase II MPEG-4 AVC 9.4. MPEG-21 Reference Software 8.10 Application formats 10.1.. Application Format Framework 10.2. PhotoPlayer Application Format 10.3 MPEG-21 Content Format 8.11 Maintenance 11.1. Systems coding standards 11.2. Video coding standards 11.3. Audio coding standards 11.4. Visual description coding standards 11.5. Audio description coding standards 11.6. MDS standards 9. Liaison matters 10. Organisation of this meeting 10.1 Tasks for subgroups 10.2 Joint meetings

22

11. Administrative matters 11.1 Schedule of future MPEG meetings 11.2 Promotional activities 12. Planning of future activities13. Resolutions of this meeting 14. A.O.B 15. Closing

Annex 3List of input documents

Number v Authors Title

12108 Wo Chang Document Register for SC29/WG11 Meeting Poznan, Poland

12109 Klaus DiepoldWo Chang AHG on MAFs

12110Sylvain DevillersIan BurnettIngo Wolf

AHG on Digital Item Streaming

12111Xin WangJaime Delgado Chris Barlas

AHG on MPEG-21 REL Profiling

12112 Teruhiko Suzuki AHG on Colour Spaces

12113 Mike KsarVladimir Levantovsky AHG on Font Format Representation

12114 Wo ChangTanya Beech AHG on Content-Based Search Framework

12115 G. SullivanA. Luthra AHG on Video IDCT Specification

12116 Yi-Shin TungChung-Neng Wang

AHG on Maintenance of MPEG-4 Visual related Documents, Reference Software and Conformance

12117

L. CieplinskiA. YamadaSang-Kyun KimRobert O'Callaghan

AHG on Maintenance of MPEG-7 Visual related Documents and Reference Software

12118 Miroslaw BoberSang-Kyun Kim

AHG on Description Tools for MPEG-7 New Visual Extensions

12119 Aljoscha SmolicHideaki Kimata AHG on 3DAV Coding

12120 Euee S. JangKohtaro Asai AHG on Video Coding Tools Repository

12121 Jizheng XuB‚atrice Pesquet-Popescu AHG on Exploration in Wavelet Video Coding

12122 Touradj EbrahimiJoern Ostermann

AHG on Organization of Workshop on Future Directions in Video Compression

12123 Robert Turney AHG on MPEG-4 Part 9 Reference Hardware

23

Marco Mattavelli Description Phase 1 and 2

12124 R. Sperschneider AHG on Audio Standards Maintenance

12125 M. Gruhne AHG on MPEG-7 Audio

12126 S. Quackenbush AHG on Spatial Audio Coding

12127 Tilman Liebchen AHG on MPEG-4 Lossless Audio Coding

12128 Paolo NesiGiorgio Zoia AHG on Symbolic Music Representation

12129 S. Quackenbush AHG On Exploration of Scalable Audio and Speech Coding

12130Marius PredaJeong-Hwan AhnFrancisco Morán

AHG on AFX documents, CEs, and software

12131Mika‰l Bourges-S‚venierVishy SwaminathanItaru Kaneko

AHG on MPEG-J GFX

12132Wo ChangAkio YamadaMiroslaw Bober

AHG on Liaison with WG1 (JPEG)

12133 Gerrard DruryFrederik De Keukaleare AHG on MPEG-21 DIP

12134 Niels Rump AHG on Functional Granularity of DII

12135 Christian TimmererThomas DeMartini AHG on MPEG-21 DIA

12136 FX NuttallAndrew Tokmakoff AHG on MPEG-21 Event Reporting

12137 Myriam AmielhSylvain Devillers AHG on MPEG 21 Fragment Identifiers

12138 Marc Gauvin AHG on RDD Implementation Issues

12139Claude SeyratAndreas HutterItaru Kaneko

AHG on XML Structure Binarization and Streaming

12140Jean-Claude DufourdAlexandre Cotarmanac.hYoung-Kwon Lim

AHG on Scene Representation

12141 David SingerVisharam Mohammed AHG on MPEG File Formats

12142Jean Gelissen HendryMartin RussDavide Rogai

AHG on MPEG Multimedia Middleware

12143M. BoberA. YamadaS. K. Kim

AHG on Photo Player MAF

12144 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 13818-4:2004/FDAM 2 [SC 29 N 6709]

12145 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-5:2001/FDAM 6 [SC 29 N 6710]

24

12146 ISMA via SC 29 Secretariat Liaison Statement from ISMA [SC 29 N 6711]

12147 TVA via SC 29 Secretariat Liaison Statement from the TV-Anytime Forum [SC 29 N 6712]

12148 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC FDIS 21000-9 [SC 29 N 6732]

12149 3GPP via SC 29 Secretariat Liaison Statement from 3GPP [SC 29 N 6742]

12150 ITU-R SG 6 via SC 29 Secretariat Liaison Statement from ITU-R SG 6/WP 6A [SC 29 N 6743]

12151 ITU-R SG 6 via SC 29 Secretariat Liaison Statement from ITU-R SG 6/WP 6A [SC 29 N 6744]

12152 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-3:2001/FDAM 6 [SC 29 N 6745]

12153 IEC TC 100 via SC 29 Secretariat IEC CDV 61966-2-4 [SC 29 N 6741]

12154 SC 29 Secretariat Summary of Voting on ISO/IEC CD 21000-17 [SC 29 N 6746]

12155 SC 29 Secretariat Summary of Voting on ISO/IEC 15938-4:2002/PDAM 2 [SC 29 N 6747]

12156 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-3:2001/FPDAM 4 [SC 29 N 6748]

12157 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-16:2004/FPDAM 1 [SC 29 N 6749]

12158 SC 29 Secretariat Summary of Voting on ISO/IEC 21000-7:2004/FPDAM 1 [SC 29 N 6751]

12159 Godfrey Rust The Model of Making in indecs and RDD

12160 Niels Rump Editors input towards DoC on PDAM1 of 21000-3

12161 Niels Rump Editors input towards FPDAM1 of 21000-3

12162 Niels Rump for the UKNB Late UKNB comment on ISO/IEC 21000-3 PDAM1

12163 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-4:2004/FDAM 6 [SC 29 N 6755]

12164 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 15938-7:2003/FDAM 1 [SC 29 N 6756]

12165 SC 37 via SC 29 Secretariat SC 37 NP on Biometric data interchange formats -- Face identity data

12166 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-3:2001/FPDAM 5 [SC 29 N 6773]

12167 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2001/FPDAM 9 [SC 29 N 6774]

12168 Jean H.A. Gelissen Multimedia Middleware (M3W) introduction

12169 Mike Ksar Updated Working Draft of Open Font Format Specification

12170 DMP via SC 29 Secretariat Liaison Statement from DMP [SC 29 N 6775]

12171 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 14496-17.2 [SC 29 N 6776]

12172 Godfrey Rust Chris Barlas

RDD Editors' comments on AhG Mandate Issues

25

Niels Rump

12173 Jean-Claude Dufourd LASeR encoding XML format

12174 Jean-Claude Dufourd LASeR reference software: codec

12175 Jean-Claude Dufourd LASeR and SAF editor's study

12176Vincent BottreauGrégoire PauJizheng Xu

Vidwav evaluation software manual

12177 Jean-Claude Dufourd Updated input device emulation proposal

12178 Jean-Claude Dufourd Compression efficiency of the BiM-compatible SDL

12179 Leonardo Chiariglione M12159 revised and commented by lc

12180Peder DregeThomas SkjølbergAndrew Perkis

MPEG-21 client and server solutions for rich mobile content delivery

12181 Alexandre Cotarmanac'h Cazoulat

Handling Video in LASER (contribution for the AhG June 27th and 28th)

12182 TVA via SC 29 Secretariat Liaison Statement from TVA [SC 29 N 6782]

12183 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 11172-6

12184 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-5:2001/FPDAM 8

12185 SC 29 Secretariat Summary of Voting on ISO/IEC 13818-1:2000/DCOR 4

12186 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-11:200X/DCOR 4

12187 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-16:2004/DCOR 2

12188 SC 29 Secretariat Summary of Voting on ISO/IEC 15938-1:2002/DCOR 2

12189 SC 29 Secretariat Summary of Voting on ISO/IEC TR 15938-8:2002/DCOR 1

12190 SC 29 Secretariat Summary of Voting on ISO/IEC CD 21000-14

12191

Thomas DeMartiniChristian TimmererBenoît PellanTruong Cong ThangEric Delfosse

Report of CE on Harmonisation of Conversion Tools

12192 Ray Taylor Editors Comments on ISO/IEC 15938-1/FPDAM2

12193 Ray Taylor Proposal for 15938-1 Amd2 conformance

12194 Thomas-Kerr (CRC-SITUoW) Ian Burnett (UoW enikos)

Bitstream Binding Language 2.0: A Response to the CfP for Digital Item Streaming

12195 Thomas-Kerr (CRC-SITUoW) Ian Burnett (UoW enikos)

Implementation System Documentation 2.0: Bitstream Binding Language

12196 HendryMunchurl Kim M3W Service Manager, Service and Metadata

12197 HendryMunchurl Kim

Protecting and Governing Music MAF Player Format based Contents by using MPEG-21 IPMP

26

12198

Yoshinori SugiharaTsuyoshi MoriyamaHideaki KimataMasaki Kitahara

Keio test data sets for MVC

12199

DevillersRansburgTimmererHellwagnerCotarmanac'h

XML Streaming Instructions: response to the CfP for Digital Item Streaming

12200

RansburgDevillersTimmererHellwagnerCotarmanac'h

Media Streaming Instructions: response to the CfP for Digital Item Streaming

12201

DevillersRansburgTimmererHellwagnerCotarmanac'h

Properties Style Sheet: response to the CfP for Digital Item Streaming

12202

Weon-Geun OhIk-Hwan ChoA-Young ChoHyun-Mi KimDong-Seok JeongHae-Kwang KimSung-Phil Heo

Feasibility Test of MPEG-7 Visual Descriptors as a visual identifier descriptor

12203

Devillers Ransburg TimmererHellwagner Cotarmanac'h

Complex-mode Streaming Instructions: response to the CfP for Digital Item Streaming

12204

Christian TimmererHermann HellwagnerAndreas HutterJörg Heuer

Signalling of buffer configuration parameters for DI Streaming: a response to the CfP on DI Streaming

12205Sylvain DevillersChristian Timmerer Joseph Thomas-Kerr

Report of CE on BSDL extensions

12206 Wo Chang White paper on CBSearch Framework

12207 Wo Chang Propose Revised of WD 1.0 Photo Player MAF

12208 Gregoire CarpentierJerome barthelemy Core experiment on weighted scales

12209 Mike Rubinfeld MPEG Vocabulary

12210 OMA via SC 29 Secretariat Liaison Statement from OMA [SC 29 N 6794]

12211 Soo-Jun Park Report of cross verification result of Image Categorisation into classes-VCE2

12212 Soo-Jun Park Report of cross verfication result of Identity-based clustering, indexing and retrieval of images- VCE-3

12213 Weongeun OhYeonjeong Jeong

MPEG-21 REL Profile for Broadcast

27

Taehyun KimHogab Kang

12214 Masaki KitaharaHideaki Kimata

Results on AVC Reference Coding of KDDI Test Sequences for Call for Proposals on Multi-View Video Coding

12215 Masaki KitaharaHideaki Kimata

Results on AVC Reference Coding of Tanimoto Laboratory Test Sequences for Call for Proposals on Multi-View Video Coding

12216

Eun-Young ChangSunyoung LeeDaiyong KimByeongwook MinNamho HurSoo In LeeEuee S. Jang

Status Report of 3DMC revision for AFX amd. 2

12217Davy De SchrijverWesley De NeveRik Van de Walle

Context-related attributes for MPEG-21 BSDL

12218

Saar De ZutterChris PoppeFrederik De KeukelaereRik Van de Walle

Contribution to WD of Amendment on C++ binding

12219 Frederik De KeukelaereRik Van de Walle

Contribution to DIP reference software - DIBO and DIXO implementations

12220

Niels RumpChris BarlasGodfrey RustSteffen Lindek

Proposed additions to the RDD to satisfy DII/Amd.1

12221 Ran BonehYotam Toib A 2x2 Upsampling Method Using Scale Prediction

12222

Paolo NesiDavide RogaiPierfrancesco BelliniJaime DelgadoEva RodríguezMaurizio Campanai

Report of CE on the indexing of IPMP protected DIDL content

12223Giorgio ZoiaPaolo NesiPierfrancesco Bellini

Draft Core Experiment methodology for MPEG-4 Symbolic Music Representation

12224

Pierfrancesco BelliniMaurizio CampanaiGiorgio ZoiaPaolo Nesi

Proposed Amendment to 14496-11 for SMR Integration and Support

12225 Werner OomenKristofer Kjörling Request for spatial audio coding profiles

12226 Werner OomenFrancois Myburg

Additional subjective test results on High quality operation modes of spatial audio coding RM0.

12227 Jeroen BreebaartChristof Faller Core experiment on non-guided spatial audio coding

12228 Werner Oomen Request for Amendment on DSD/DST support in

28

Jan van der Meer MPEG-2 Systems

12229Jaime DelgadoJose PradosEva Rodriguez

An MPEG-21 REL mobile profile

12230

Alexander Behrens Emin Martinian Huifang Sun Anthony Vetro Jun Xin

Comments on Call for Proposals on Multiview Video Coding

12231

Alexander Behrens Emin Martinian Huifang Sun Anthony Vetro Jun Xin

Multiview Video Anchor Coding Results for Ballroom and Exit

12232 David Singer Considerations for the SVC File format

12233 David Singer Corrigendum issues for the ISO File format

12234 Gwang Hoon ParkKyung Il Lee

Advanced FMO for ROI enhancement of the cyclic-FGS-based video coding

12235Sang-Ki KimSangyoun LeeSun-Jeong Kim

Propose Thermal Image Sensor for VCE-4

12236Hyoung-Joong KimYong-Soo ChoiYong-Ju Cho

The Extended Symbolic Music Representation for Korean Traditional Music.

12237 Mikael Bourges-Sevenier AhG on GFX, Tokyo meeting report

12238 Mikael Bourges-SevenierVishy Swaminathan Draft updated ISO/IEC 14496-21 CD study

12239 Hideaki KimataMasaki Kitahara

Usage and description of camera parameters for Multi-view Video Coding and Free-viewpoint TV

12240Yung-Lyul LeeYung-Ki LeeDae-Hee Kim

Anchor coding results of HHI sequences for Call for Proposal on Multi-view video coding

12241 Holger CrysandtMarkus Beermann Cross check results of the EnhancedAudioSignature

12242

Kazuo SugimotoYoshihisa YamadaKohtaro AsaiTokumichi Murakami

Contribution to VCTR Reference Software - Intra Only Decoder for AVC Baseline Profile

12243

Kazuo SugimotoYoshihisa YamadaKohtaro AsaiTokumichi Murakami

Proposed update of TD 3.0

12244Yutaka KamamotoNoboru HaradaTakehiro Moriya

Proposed corrections to avoid inconsistency in study on FPDAM4 ALS

12245 Jens-Rainer OhmGary Sullivan Introduction to MPEG-1 Video (draft)

12246 Jens-Rainer OhmGary Sullivan Introduction to MPEG-2 Video (draft)

29

12247 Jens-Rainer OhmGary Sullivan Introduction to MPEG-4 Visual (draft)

12248 Jens-Rainer OhmGary Sullivan Introduction to MPEG-4 AVC (draft)

12249 Jens-Rainer OhmGary Sullivan Introduction to MPEG Video Explorations (draft)

12250

Satoshi ItoToru KambayashiHideyuki AisuHiroshi Niina

Use case scenarios and REL Profile requirements for pre-recorded optical media

12251 Teruhiko Suzuki Requirements of new colour spaces for MPEG video

12252 Fons Bruls Results on AVC Reference Coding of Tanimoto Laboratory Test Sequences for 3DAV CfP

12253 Kevin M. Short KOZ scalable audio compression

12254 Schuyler Quackenbush 72nd MPEG Audio Report

12255Sangyoun LeeSun-Jeong KimJongheum Kim

Report of cross verification result - VCE-2

12256Sangyoun LeeJung-Sup RyuSegon Jang

Report of cross verification result- VCE-3

12257 Marc GauvinJose Neri sDae on RDD Implementation Issues AHG

12258Gregoire Pau Philippe de Cuetos Claude Seyrat

MPEG-21 binarisation status with BiM reference software

12259 Maaskant (Philips) Hendry (ICU) Martin Russ (BT) Davide Rogai (UNIFI- DISIT) Updated version of the M3W WD

12260 Maaskant (Philips) Jean Gelissen (Philips) Reply to the M3W extended call for proposals

12261 Melanie Dulong de Rosnay Contribution to AhG on MPEG-21 RDD

12262 Jin LiJames D. Johnston

A Progressive to Lossless Embedded Audio Coder (PLEAC)

12263 A. G. Tescher for USNB USNB Contribution: MPEG-4 ALS issues

12264 A. G. Tescher for USNB USNB Contribution: Complexity issues regarding Audio tools

12265

In Yong ChoiSang Bae ChonJeongil SeoInseon JangKoeng-Mo SungKyeongok Kang

Objective Measurement of Total Auditory Quality of Spatial Audio Coding

12266

Jeongil SeoSeungkwon BeackIn Yong ChoiKyeongok KangJinwoo HongMinsoo Hahn

An Efficient Quantization Method for CLD

12267 Chong Kok Seng Request for formal verification test of the Low

30

Takeshi Norimatsu Yoshiaki Takagi Neo Sua Hong

Complexity mode of Spatial Audio Coding RM0 and backward compatibility with LP-SBR

12268

Hee-Suk PangDong Soo KimJaehyun LimHYEN-O OH

Subband domain energy compensation for spatial audio coding

12269 CJ Lee for SGNB SGNB comment on request to extend the work schedule for Spatial Audio Coding

12270

Hyok SongByeongho ChoiJinwoo BaeJisang Yoo

A framework for Multi-view Video Coding using Accumulated image

12271 Kyuheon Kim Korean NB's Comment on Spatial Audio Coding

12272 Myriam AmielhSylvain Devillers Editor's input on ISO/IEC 21000-17 (FID)

12273 Myriam Amielh CE report on Logical Model Description

12274

Sang-Kyun KimSeungji YangJun Ho ChoYong Man RoJi-Yeon Kim

A MAF Photo Player for Enhanced Photo Browsing

12275

Sang-Kyun KimSeungji YangKyong Sok SeoYong Man RoJi-Yeon Kim

CE Report for Photo Categorization (VCE-2)

12276

Sang-Kyun KimKyong Sok SeoSeungji YangYong Man RoJi-Yeon Kim

Test Dataset for Person-identity-based Photo Clustering Experiment (VCE-3)

12277

Sang-Kyun KimSeungji YangKyong Sok SeoYong Man RoJi-Yeon Kim

CE Report for Person-identity-based Photo Clustering Experiment (VCE-3)

12278

Yo-Sung HoSeung-Uk YoonEun-Kyung LeeSung-Yeol KimSeung-Hwan KimKugjin YunDaehee Kim

Intermediate Result on Multi-view Video Coding using Layered Depth Images

12279Pierfrancesco BelliniPaolo NesiMaurizio Campanai

MPEG-4 SMR RM0 documentation for SMR WD text

12280Daehee KimNamho HurSoo In Lee

Anchor bitstreams for Call for Proposals on multi-view video coding

12281 Philippe de Cuetos LASeR: Improvement of compression performances

31

Cedric Thienot Robin Berjon Claude Seyrat

of BiM-Compatible SimpleEncoding

12282Jeong Taek Hwang Doug Young SuhYongju Cho

Proposal for storage and transport of SVC streams in the non-sequential order

12283 Akio YamadaSang-Kyun Kim

MPEG-7 Visual part of eXperimentation Model Version 24.1

12284 Akio YamadaSang-Kyun Kim WD 2.1 of ISO/IEC TR 15938-8:2003/Amd.2

12285 Akio Yamada A Preliminary Study on the addition of a file format for Image Player

12286Ruiqin XiongJizheng XuFeng Wu

Responses to Vidwav EE1

12287 Philippe de Cuetos Claude Seyrat Editor's Study of ISO/IEC 23001-1 (BiM)

12288

Jeong-Hwan AhnMahnjin HanSeong Won Ryuon behalf of Mobile 3D Standardization Forum in Korea

Results on Core Experiment of 3D Compression Stream Design

12289Zhongyang HuangShengmei ShenTakafumi Ueno

IPMP Protection for MPEG-A Music Player

12290 Akio YamadaSang-Kyun Kim Editors' Study of ISO/IEC 15938-3:2002/FPDAM 2

12291

Jeong-Hwan AhnMahnjin HanSeong Won Ryuon behalf of Mobile 3D Standardization Forum in Korea

Conformance Test of Simple 3D Compression Profiles and Levels

12292

Ihab AmerMohammed SayedWael BadawyGraham Jullien

A SystemC Modeling Strategy to Integrate Selected IP-Blocks with the H.264 Software Reference Model

12293

Jeong-Hwan AhnMahnjin HanSeong Won Ryuon behalf of Mobile 3D Standardization Forum in Korea

Proposal of 3D Compression Profile and Level Indication

12294 Yifeng QiuWael Badawy

Updated Hardware Blocks and Conformance Tests for the Multiple IP-Core Hardware-Accelerated Software System Framework for MPEG4

12295 Yifeng QiuWael Badawy

An Integrated Virtual Socket Hardware-Accelerated Software Framework for MPEG4

12296 Akio YamadaR. O'Callaghan Editors's Study of ISO/IEC 15938-8:2002/DCOR 1

12297 Jaime DelgadoJose Prados

Comparison of REL mobile profiles

32

Eva Rodriguez

12298Aljoscha Smolic Philipp Merkle Karsten Mueller

Anchor Coding Results for Fraunhofer HHI Test Sequences

12299Jaime DelgadoEva RodriguezMaria Cirera

Core Experiment on use of Event Report Requests: Specification of Use Cases

12300

Matthias Kautzner Karsten Mueller Aljoscha Smolic Thomas Wiegand

Preliminary Results on EE1 for D3DMC

12301

Philipp Merkle Aljoscha Smolic Karsten Mueller Thomas Wiegand

Statistical Evaluation of Spatiotemporal Prediction for MVC

12302 Eunmi OhYang-Seock Seo

Response to CfI on scalable speech and audio coding: The need for a new coding paradigm and its application scenarios

12303 Grégoire Pau Maria TrocanBéatrice Pesquet-Popescu

Bidirectional Joint Motion Estimation for Vidwav Software

12304

Isabelle AmonouNathalie Cammas Sylvain Kervadec Stephane Pateux

France Telecom response to CE5 on quality layers

12305

Isabelle Amonou Nathalie Cammas Sylvain Kervadec Stephane Pateux

France Telecom response to CE9 on improved quantization

12306

Isabelle Amonou Nathalie Cammas Sylvain Kervadec Stephane Pateux

Syntax for FGS pass fractionning

12307

Isabelle Amonou Nathalie Cammas Sylvain Kervadec Stephane Pateux

On the high level syntax for SVC

12308

Isabelle Amonou Nathalie Cammas Sylvain Kervadec Stephane Pateux

Cross check of Nokia's proposal for CE5 on quality layers

12309

Isabelle Amonou Nathalie Cammas Sylvain Kervadec Stephane Pateux

Cross check of RWTH's proposal for CE2 on deblocking filter

12310

Jongheum KimSungphil HeoWeongeun OhHaekwang Kim

IPMP metadata protection for super-distribution MAF

12311 Anisse TalebScalable Speech and Audio Coding based on the 3GPP AMR-WB+ standard: Response to the CfI on scalable speech and audio coding

33

12312 Karol WnukowiczWladyslaw Skarbek

Dominant Color Temperatures - a new descriptor or an application of the Dominant Color descriptor

12313 MenHuang Lee SVC CE1 - Verification of VisioWave's proposal for CE1

12314 MenHuang Lee SVC CE6 - Verification of MSRA's proposal for CE6

12315

MenHuang LeeChinPhek OngChongSoon LimShengMei ShenSatoshi KondoYoshimasa Honda

SVC CE1 - Reduced Memory MCTF

12316

ChinPhek OngChongSoon LimMenHuang LeeShengMei ShenYoshimasa HondaSatoshi Kondo

SVC CE1 - Picture Duplication Process

12317 - -

12318

MenHuang LeeChinPhek OngChongSoon LimShengMei ShenYoshimasa HondaSatoshi Kondo

SVC - Progressive MCTF

12319Kristofer KjörlingWerner OomenJürgen Herre

MPEG Surround, a proposed new name for the Spatial Audio Coding tool

12320

Kristofer KjörlingWerner OomenFrancois MyburgJürgen Herre

Proposed corrections and clarifications to the WD on Spatial Audio Coding

12321

Truong Cong ThangTae Meon BaeYong Ju JungYong Man RoJung Won KangHaechul ChoiJae-Gon Kim

SVC CE8: Spatial scalability of multiple ROIs

12322 Honggang Qi Wen Gao Debin Zhao Siwei Ma Liang Wang Long Xu

High Accurate and multiplierless fixed-point DCT for CFP on 8x8 fixed-point IDCT and DCT standards

12323

Truong Cong ThangTae Meon BaeDuck Yeon KimYong Man RoJung Won KangJae-Gon KimJin-Woo Hong

FMO implementation in JSVM

12324 Honggang Qi Wen Gao Debin Zhao Siwei Ma Liang Wang Long Xu

An example of fixed-point IDCT for CFP on fixed-point 8x8 IDCT and DCT standard

34

12325 Tilman LiebchenProposed Text of ISO/IEC 14496-3:2001/FDAM 4, Audio Lossless Coding (ALS), new audio profiles and BSAC extensions

12326 Tilman Liebchen Proposed Corrections and Modifications of Study on FPDAM4 (ALS)

12327 Tilman Liebchen Additional Information on the Performance of ALS

12328

Kwanghoon SohnYongtae KimChangseob ParkJaeho Lee

Comments on requirement and camera parameters

12329 Marcos AvilésFrancisco Morán Spatial and SNR scalable coding of WSSs

12330 Leszek CieplinskiRobert O'Callaghan

UK NB comments on the text of ISO/IEC 15938-3:2002/FPDAM 2

12331 Robert O'CallaghanMiroslaw Bober A Face Recognition Dataset from Video Content

12332Michael RansburgChristian TimmererHermann Hellwagner

Proposed amendments to 14496-12 and/or 21000-9 for storage of timed metadata

12333

Michael RansburgEric DelfosseChristian TimmererHermann Hellwagner

Proposed DIA extensions: AdaptationQoS for DI Streaming

12334 Haechul ChoiNam-Hyeong Kim

Embedded quantization for improved SNR scalability

12335Zvi LifshitzCyril ConcolatoTIRAMISU consortium

Proposal for free-distribution MAF

12336 Haibin Huang Wee Boon Choo

Proposed Corrections to Study on MPEG-4 ALS FPDAM4

12337Rongshan Yu Xiao Lin Susanto Rahardja

Integration of MPEG-4 SLS with MPEG-4 scalable audio coding tools

12338

Masayuki TanimotoToshiaki FujiiTakanori SenohTerumasa AokiYoshinori Sugihara

Test Sequences with Different Camera Arrangements for Call for Proposals on Multiview Video Coding

12339

Xiong Xiangyang Ji Dongdong Zhang Jizheng Xu (Microsoft Research Asia) Grégoire Pau Maria Trocan (ENST) Vincent Bottreau (INRIA)

Vidwav Wavelet Video Coding Specifications

12340 Cyril ConcolatoJean Le Feuvre

Editorial and technical inputs for LASeR study of FCD

12341

Hyon-Gon ChooGun BangBumsuk ChoiYoungBae ByunJeho Nam

IPMP Interfaces of Multimedia API on M3W

35

12342 Vincent Bottreau Cross verification report of France Telecom Division R&D response to CE9 on improved quantization

12343 Vincent Bottreau CABAC variant and options for bit error resilience

12344 K GrantP Jessop UK National Body Position on Meeting Accessibility

12345Thomas RathgenPeter AmonAndreas Hutter

Proposal for SVC File Format Specification

12346 Andreas Schneider proposed study on Parametric Stereo Conformance

12347 Schuyler Quackenbush Proposed Tutorial on MPEG Surround Audio Coding

12348 Pierrick Philippe for FNB FNB Contribution: Spatial Audio Coding Activity issues

12349 Pierrick Philippe for FNB FNB Contribution : Comments on N7016 (signaling and decoding behavior)

12350David ViretteManuel BriandPierrick Philippe

Comments on the Spatial Audio Coding RM0 and Working Draft

12351 Vincent Bottreau FGS bitplane coding

12352 Andreas Hutter Comments on ISO/IEC 15938-1:2002/FPDAM 2

12353Herbert Thoma Stefan Krägeloh Harald Fuchs

Application Scenarios and Derived Requirements for Video Player MAF

12354Julien ReichelDiego Santa-CruzFrancesco Ziliani

SVC CE1-Improved Residual Picture Buffer Management

12355

Marco TagliasacchiStefano TubaroJoao AscensoKannan RamchandranRohit PuriAb

Distributed video coding: a functional analysis

12356Jaime DelgadoEva RodriguezMaria Cirera

Core Experiment on use of Event Report Requests: Software Implementation

12357Qian Hu QingXiao Jiang Lu Yu

Hardware Implementation of AVS 8x8 Inverse Integer Cosine Transform Module

12358Jaime DelgadoRoberto GarciaEva Rodriguez

From Abstraction to Manifestation

12359 Ci-Xun ZhangLu Yu Fixed-Point 8x8 IDCT

12360 Stefan KraegelohHarald Fuchs Proposed Protected Music Player MAF

12361 Kristofer Kjörling Proposal for extended signalling in Spatial Audio

12362 Michael ZuffereyEric Delfosse

Proposed DIA extensions: Multiple input/output values for AdaptationQoS IOPins and MPEG-7 as

36

Christian TimmererHarald Kosch BSDLink steering description

12363Christian TimmererMichael RansburgHermann Hellwagner

Comments on DI streaming requirements

12364 Christian Timmerer Editors' input on Study of ISO/IEC 21000-7 FPDAM/1

12365

Cristina GomilaYeping SuAntonio OrtegaJae Hoon Kim

Comments on the Preliminary CfP on Multi-view Video Coding

12366

Xin WangThomas DeMartiniMai NguyenManuel Ham

An Improvement of the MPEG REL Profile under Consideration in Response to the Requirements for Pre-recorded Optical Media

12367 Matthias Gruhne CE Crosscheck on Weighted Scale Type

12368

A. HoelzerW. OomenK. KjörlingJ. Herre

Additional subjective test results on R/D scalability of Spatial Audio Coding RM0

12369

S.DischM. NeusingerG. HothoJ. Herre

Proposed Core Experiment on Guided Envelope Shaping for Spatial Audio Coding

12370 Ye-Kui WangMiska M. Hannuksela Proposed SVC file format working draft

12371Alberto CastilloFrancisco MoránJosé M. Martínez

A First Draft for Fractalish Nature Description Schemes

12372 Ralph SperschneiderManfred Lutzky

Proposed new profile for high-quality communication

12373Feng-Cheng ChangChien-Peng HoChun-Jen Tsai

Update to the FGS-Based Multimedia Resource Delivery Test Bed Software

12374 Markus Beermann Joint reduction of ringing and blocking for VidWav

12375 Marius Preda Francoise Preteux

MPEG-4 SDK : beyond MPEG-4 Implementation Software

12376 Yongjun WuJohn Woods

Aliasing reduction for subband/wavelet scalable video coding

12377

A. SalomieR. DeklerckD. CostinA. MarkovaA. MunteanuP. Schelkens

Updates to MeshGrid

12378 Mohammed Zubair VisharamAli Tabatabai Proposed WD for the SVC File Format

12379 Rongshan YuRalf Geiger

Proposed Text of ISO/IEC 14496-3:2001/FDAM 5, Scalable Lossless Coding (SLS)

37

Xiao LinSusanto Rahardja

12380

Rongshan Yu Ralf Geiger Xiao Lin Susanto Rahardja

Update on Performance and Complexity of Lossless Audio Coding Architectures

12381

Ralf Geiger Markus Schmidt Jürgen Herre Thomas Sporer

Proposed Verification Test Methods for MPEG-4 SLS

12382 Jean-Claude Dufourd Updated LASeR mini and full profile definitions

12383 Jean-Claude Dufourd LASeR conformance streams

12384 Jean-Claude Dufourd French NB comment on LASeR

12385 Jean-Claude Dufourd LASeR utility software

12386 Jean-Claude Dufourd Updated proposals for LASeR Technologies under consideration

12387 Gary J. Sullivan Prop. Changes to 11172-6 MPEG-1 Video IDCT precision spec. FCD

12388 Gary J. Sullivan Prop. Corrigendum to 11172-2 MPEG-1 Video for IDCT issues

12389 Gary J. Sullivan Prop. Corrigendum to Rec. H.262 / 13818-2 MPEG-2 Video for IDCT issues

12390 Gary J. SullivanJens-Rainer Ohm

Prop. Corrigendum to 14496-2 MPEG-4 Visual for IDCT, FGS, and OBMC issues

12391Do-Hyung KimSang-Wook KimMiyoung Kim

Proposed Integration of BSAC and SBR

12392Miyoung KimSang-Wook KimDo-Hyung Kim

Proposed New Audio Conformance Bitstreams for BSAC

12393Sang-Wook KimMiyoung KimDo-Hyung Kim

Proposed Study on 14496-3:2005/DCOR1:2005

12394 Kyuheon Kim for KNB Korean NB Contribution: Public MPEG Software issues

12395

YeSun JoungYoung-kwon LimWon-sik CheongJihun ChaKyuheon Kim

Comments and considerations on LASeR Study of FCD

12396 Jeongyeon LimMunchurl Kim

Synchronization of Multiple JPEG data to MP3 tracks in Music MAF Player Format

12397 Gyeong Ja Jang Shinjun Lee Keun Ho Kim Mahnjin Han

Experiment for the update of Depth Image-based Representation

12398 SC 29 Secretariat Liaison Statement from ECMA

12399 Sunyoung LeeHyunsoo AhnHyungyu Kim

Reconfigurable Syntax Description for VCTR

38

Chungku YieEuee S. Jang

12400

Hyungyu KimSunyoung LeeHyunsoo AhnChungku YieEuee S. Jang

Analysis of Python Language for VCTR Implementation

12401

Hyunsoo AhnSunyoung LeeHyungyu KimChungku YieEuee S. Jang

Status Report of VCTR Reference Software

12402 Akio Ishikawa KDDI Test Sequences for Call for Proposals on Multi-View Video Coding

12403Callow (HI Corporation)Kaneko (Tokyo Polytechnic University)Sone (Tokyo Polytechnic University)

Portable Music player with 3D animation using ISO/IEC 14496-21 CD (MPEG-J GFX)

12404Stephen DavisIan BurnettJoseph Thomas-Kerr

FRUs - Suggested Text for Fragment Request Units

12405 Raphaèle BalterPatrick Gioia Conformance bitstreams for the AFX 3D profile

Annex 4List of output documents

No. Source Title

7264 Convener List of Documents from the Poznan, Poland Meeting

7265 Convener Resolutions of the Poznan, Poland

7266 Convener List of AHGs Established at the 73rd Meeting in Poznan, Poland

7267 Convener Report of the 73rd Meeting in Poznan, Poland

7268 Convener Guidelines for Electronic Distribution of MPEG and WG 11 Documents

7269 Convener Press Release of the 73rd Meeting in Poznan, Poland

7270 Convener Meeting Notice of the 74th Meeting in Nice, France

7271 HoD Guide for WG 11 Meeting Hosts

7272 HoD MPEG 101

7273 ISG Status of HDL submissions and commitments for MPEG-4 Part-9

7274 ISG AHG on MPEG-4 Part 9: Reference Hardware Description Phase 1 and 2.

7275 ISG Study of Part 9 Reference Hardware Description

7276 Requirements MP4 Profiles under Consideration

7277 Requirements MP21 Profiles under Consideration

39

7278 Requirements DI Streaming Requirements

7279 Requirements Requirements on XML Fragment Requests

7280 Requirements MAFs under Consideration

7281 Requirements MPEG-A Part 1 PDTR

7282 Requirements Requirements on Multi-view Video Coding

7283 Requirements Report of Activities on Supporting New Colour Spaces in MPEG Standards

7284 Convener AHG on MPEG Dictionary

7285 Convener AHG on MAFs

7286 Convener AHG on Content Based Search Framework

7287 Convener AHG on XML Fragment Requests

7288 Convener AHG on Colour Spaces

7289 Video Text of ISO/IEC 11172-2 DCOR4

7290 Video Introduction to MPEG-1 Video

7291 Video Text of ISO/IEC 11172-4 DCOR1

7292 Video Study of ISO/IEC 11172-6 FCD

7293 Video Text of ISO/IEC 13818-2:2000 DCOR2

7294 Video Introduction to MPEG-2 Video

7295 Video Text of ISO/IEC 13818-4:2004 DCOR1

7296 Video Text of ISO/IEC 14496-2:2004 DCOR2

7297 Video Introduction to MPEG-4 Part 2 Video (rectangular)

7298 Video Introduction to MPEG-4 Part 2 Video (arbitrary shape)

7299 Video Text of ISO/IEC 14496-4:2004 DCOR2

7300 Convener Terms of Reference

7301 Convener MPEG Standards

7302 Convener Table of unpublished standards at FDIS level

7303 Convener Work plan and time line

7304 Convener Work item assignment

7305 Convener List of patent statements received

7306 Video Disposition of Comments on ISO/IEC 14496-4:2004 FPDAM9

7307 Video Text of ISO/IEC 14496-4:2004 FDAM9

7308 Video Disposition of Comments on ISO/IEC 14496-5:2001 FPDAM8

7309 Video Text of ISO/IEC 14496-5:2001 FDAM8

7310 Video Working Draft 3 of ISO/IEC 14496-10:200x/AMD1 Scalable Video Coding

7311 Video Joint Scalable Video Model (JSVM) 3

7312 Video JSVM 3 Software

40

7313 Video Justification of New 4:4:4 Profiles

7314 Video Introduction to MPEG-4 Part 10 Advanced Video Coding

7315 Video Introduction to SVC Extension of Advanced Video Coding

7316 Video Study of ISO/IEC 15938-3:2002/FPDAM2

7317 Video MPEG-7 Visual XM Document version 25.0

7318 Video Description of Core Experiments for MPEG-7 New Visual Extensions

7319 Video Introduction to Low-level Visual Description Tools

7320 Video Introduction to High-level Visual Description Tools

7321 Video Text of ISO/IEC TR 15938-8:2002 COR1

7322 Video Text of ISO/IEC TR 15938-8:2002 DAM2

7323 Video Request for Subdivision of ISO/IEC 23000: Part 3 Photo Player Multimedia Application Format

7324 Video WD 2.0 of 23000-3 Photo Player Multimedia Application Format

7325 Video Introduction to Photo Player Multimedia Application Format

7326 Video Introduction to Accuracy Requirements for 8x8 IDCT

7327 Video Call for Proposals on Multi-View Video Coding

7328 Video Introduction to Multi-View Video Coding

7329 Video Study of Video Coding Tools Repository V5.0

7330 Video VCTR Textual Description V4.0

7331 Video VCTR Software V3.0

7332 Video Introduction to Video Coding Tools Repository

7333 Video Description of Exploration Experiments in Wavelet Video Coding

7334 Video Wavelet Codec Reference Document and Software Manual V1.0

7335 Video Call for Proposals on Fixed-point 8x8 IDCT and DCT

7336 Video Introduction to Fixed-point Implementation of DCT/IDCT

7337 Video Second Workshop on Future Directions in Video Compression – Call for Participation

7338 Convener AHG on Video IDCT Specification

7339 Convener AHG on Maintenance of MPEG-4 Visual related Documents, Reference Software and Conformance

7340 Convener AHG on Maintenance of MPEG-7 Visual related Documents and Reference Software

7341 Convener AHG on Description Tools for MPEG-7 New Visual Extensions

7342 Convener AHG on 3D Video Coding

7343 Convener AHG on Video Coding Tools Repository

7344 Convener AHG on Exploration in Wavelet Video Coding

7345 Convener AHG on Organization of Workshop on Future Directions in Video

41

Compression

7346 Convenor AHG on Photo Player MAF

7347 Integration WD2.0 of ISO/IEC 14496-4:200x/ AMD12 Morphing & Textures

7348 Integration WD2.0 of ISO/IEC 14496-5:200x/AMD9 Morphing & Textures

7349 Liaison Responses to National Body Comments

7350 Liaison List of Organisations with which MPEG entertains liaisons (as of July 2005)

7351 Liaison Liaison Statement to ITU-R SG 6/WP 6A on ITU-R BS.1387-1

7352 Liaison Liaison Statement to ITU-R SG 6 WP 6A on objective evaluation of spatial audio reproduction

7353 Liaison Liaison Statement to SMPTE on OpenType® and ISO/IEC 14496-22

7354 Liaison Liaison Statement to ECMA on Universal 3D File Format

7355 Liaison Liaison Statement to JTC1 SC 37

7356 Liaison Liaison Statement to 3GPP

7357 Liaison Liaison Statement to IEC TC100

7358 Liaison Liaison Statement to TVAF

7359 Liaison Liaison Statement to OMA

7360 Liaison Liaison Statement to DMP

7361 Liaison Liaison Statement to TC 42 WG18

7362 Liaison Liaison Statement to Khronos OpenMax

7363 Audio DoC of 14496-3:2001/FPDAM 4, Audio Lossless Coding (ALS), new audio profiles and BSAC extensions

7364 Audio Text of 14496-3:2001/FDAM 4, Audio Lossless Coding (ALS), new audio profiles and BSAC extensions

7365 Audio DoC of 14496-3:2001/FPDAM 5, Scalable Lossless Coding (SLS)

7366 Audio Text of 14496-3:2001/FDAM 5, Scalable Lossless Coding (SLS)

7367 Audio Request for Amendment 14496-3:2005/AMD 1

7368 Audio Text of 14496-3:2005/PDAM 1, Low Delay AAC Profile

7369 Audio Study on 14496-3:2005/DCOR 1, BSAC and LATM

7370 Audio Text of 14496-3:2005/DCOR 2, SSC

7371 Audio Draft Performance and Complexity of MPEG Lossless Audio Coding Standards

7372 Audio Workplan for Lossless Coding

7373 Audio WD on Integration of BSAC and SBR

7374 Audio Workplan on BSAC multichannel reference software

7375 Audio Core Experiment Methodology for Symbolic Music Representation

7376 Audio Workplan for SMR Core Experiment on Traditional Korean Music

7377 Audio Workplan for SMR Core Experiment on Braille and Spoken Music

42

7378 Audio Text of Working Draft for Symbolic Music Representation (SMR)

7379 Audio Symbolic Music Representation in MPEG, an overview

7380 Audio Study on 14496-4:2004/FPDAM11, Parametric Stereo Conformance

7381 Audio Proposed correction to Audio part of 14496-4:2004

7382 Audio WD 2 of conformance of MPEG-1/2 Audio in MPEG-4

7383 Audio Status of MPEG-4 Audio Conformance

7384 Audio DoC on 15938-4:2002/PDAM 4 (Audio High-Level Description Extensions)

7385 Audio Text of 15938-4:2002/FPAM 4 (Audio High-Level Description Extensions)

7386 Audio Workplan for MPEG-7 Audio

7387 Audio WD 2 for MPEG Surround

7388 Audio Performance of MPEG Surround RM Technology

7389 Audio Workplan for MPEG Surround

7390 Audio Tutorial on MPEG Surround

7391 Audio Study on Profiles for MPEG Surround

7392 Audio Brief Report on Call for Information on Scalable Speech and Audio Coding

7393 Convener AHG on Audio Standards Maintenance

7394 Convener AHG on MPEG-7 Audio

7395 Convener AHG on MPEG Surround

7396 Convener AHG on MPEG-4 Lossless Audio Coding

7397 Convener AHG on Symbolic Music Representation

7398 Convener AHG On Exploration of Scalable Audio and Speech Coding

7399 SNHC DoC on ISO/IEC 14496-16/FPDAM1

7400 SNHC Text of ISO/IEC 14496-16/FDAM1 (Morphing & Textures)

7401 SNHC WD 2.0 of 14496-16 2nd edition

7402 SNHC WD 2.0 of 14496-16/AMD2 (Geometry & Shadow)

7403 SNHC Text of ISO/IEC 14496-16:2004/COR2

7404 SNHC Text of ISO/IEC 14496-16:2004/DCOR3

7405 SNHC AFX Core Experiments Description

7406 SNHC SNHC FAQ 14.0

7407 SNHC DoC on ISO/IEC 14496-21/CD

7408 SNHC Text of ISO/IEC 14496-21/FCD

7409 SNHC MPEG-J GFX white paper

7410 SNHC MPEG-J white paper

7411 SNHC Ad Hoc Group on AFX documents, CEs, and software

7412 SNHC AHG on MPEG-J Graphics Framework eXtensions

43

7413 Liaison Liaison Statement to DVB TM-AVC

7414 Liaison Liaison Statement to ATSC S6

7415 Liaison Liaison Statement to SCTE DVS

7416 Liaison Liaison Statement to DVD Forum WG-1

7417 Liaison Liaison Statement to Blu-ray Disc Association TEG-2

7418 MDS Introduction to MPEG-7 MDS – an overview

7419 MDS Schema Files for MPEG-21 Standards

7420 MDS Introducing ISO/IEC 21000 Schema Files

7421 MDS MPEG-21 Workshop presentations

7422 MDS Introducing MPEG-21 Digital Item Declaration

7423 MDS Introducing ISO/IEC 21000-3 DII

7424 MDS DoC of PDAM/1 on DII Related Identifiers

7425 MDS Study of PDAM/1 on DII Related Identifiers

7426 MDS Study of ISO/IEC 21000-4 IPMP Components FCD

7427 MDS Introducing the MPEG REL

7428 MDS Request for Amendment 1 of ISO/IEC 21000-5

7429 MDS PDAM/1 MPEG-21 REL Profiles

7430 MDS Request for Amendment 1 of ISO/IEC 21000-6

7431 MDS PDAM/1 ISO/IEC21000-6 Digital Item Identifier Relationship Types

7432 MDS Workplan for CE on RDD Abstraction and Resource

7433 MDS Introducing the MPEG-21 RDD

7434 MDS Study of ISO/IEC 21000-7 FPDAM/1

7435 MDS TuC for Potential 21000-7 MPEG-21 DIA Amendment(s) v3.0

7436 MDS WD of ISO/IEC 21000-7 AMD/2 v1.0

7437 MDS DIA Reference Software Status and Workplan

7438 MDS Workplan for CE on Adaptation QoS and BSDLink extensions

7439 MDS Workplan for CE on BSDL Extensions

7440 MDS Introducing the MPEG-21 Reference Software

7441 MDS Request for Amendment 1 of ISO/IEC 21000-10

7442 MDS PDAM/1 of ISO/IEC21000-10 DIP: Additional C++ Bindings

7443 MDS DoC on ISO/IEC 21000-15 CD MPEG-21 ER

7444 MDS FCD of ISO/IEC 21000-15 MPEG-21 ER

7445 MDS DoC on ISO/IEC 21000-17 CD MPEG-21 Fragment Identification of MPEG Resources

7446 MDS ISO/IEC 21000-17 FCD MPEG-21 Fragment Identification of MPEG Resources

44

7447 MDS Working Draft

7448 MDS Workplan for CE on Fragmentation for DI Streaming

7449 MDS Workplan for CE on DI Mapping to Delivery channels

7450 Convener AHG on MPEG-21 DIP

7451 Convener AHG on MPEG-21 DIA

7452 Convener AHG on MPEG-21 Event Reporting

7453 Convener AHG on MPEG-21 IPMP Components

7454 Convener AHG on MPEG-21 DIStreaming

7455 Requirements Overview of ISO/IEC TR 21000-11 Evaluation Methods for PersistentAssociation Technologies (PAT)

7456 SNHC Overview of FBA (Face and Body Animation)

7457 SNHC Overview of 3DMC (3D Mesh Coding)

7458 SNHC Overview of AFX (Animation Framework eXtension)

7459 Liaison Liaison to 3GPP on loss resilience video experiments

7460 Integration Request for ISO/IEC 14496-4:2004 AMD13 Parametric Coding for High Quality Audio Conformance

7461 Integration Text of ISO/IEC 14496-4:2004 PDAM13 Parametric Coding for High Quality Audio Conformance

7462 Integration Request for ISO/IEC 14496-5:200x AMD10 Parametric Coding for High Quality Audio Reference Software

7463 Integration Text of ISO/IEC 14496-5:200x PDAM10 SSC, DST, ALS, SLS Reference Software

7464 Audio Description of Spectral and Replication

7465 Audio Description of Lossless Coding of Over Sample Audio

7466 Audio Description of MPEG-7 Low level descriptors

7467 Audio Description of MPEG-7 high level descriptors

7468 Audio Description of Parametric Coding of High Quality Audio

7469 Systems Text of ISO/IEC 13818-1:2000/COR4

7470 Systems Request of Amendment of ISO/IEC 13818-1:2000

7471 Systems Text of ISO/IEC 13818-1:2000/PDAM 6

7472 Systems Text of ISO/IEC 14496-1:2004/PDAM2 3D Compression Profile Descriptor

7473 Systems Text of ISO/IEC 14496-11:2004/COR4 Node Coding Table Cor.

7474 Systems Text of ISO/IEC 14496-12/DCOR2

7475 Systems WD of ISO/IEC 14496-12/Amd/2

7476 Systems Text of ISO/IEC 14496-15/DCOR1

7477 Systems VM1.0 of ISO/IEC 14496-15/Amd.2 (SVC File Format)

7478 Systems DoC on ISO/IEC 14496-17/FCD

45

7479 Systems Text of ISO/IEC 14496-17/FDIS

7480 Systems Study Text of ISO/IEC 14496-20/FCD

7481 Systems TuC for ISO/IEC 14496-20

7482 Systems Validation Experiment on LASeR Binary Encoding

7483 Systems Validation Experiment on LASeR streamSource

7484 Systems Request for a new part of ISO/IEC 14496

7485 Systems Text of ISO/IEC 14496-22/CD

7486 Systems Study Text on ISO/IEC 15938-1/FPDAM2

7487 Systems WD 1.0 on MPEG-7 Systems Amd.2 Conformance (Fast Access Extension)

7488 Systems MPEG-7 Systems reference software workplan

7489 Systems DoC on Text of ISO/IEC 15938-1/DCOR2

7490 Systems Text of ISO/IEC 15938-1/COR2

7491 Systems Study Text on ISO/IEC 23001-1/FCD

7492 Systems Request for a new project on MPEG Multimedia Middleware

7493 Systems Evaluation report on submission to M3W Extended Call for Proposal

7494 Systems WD2.0 of ISO/IEC 23004-1 Architecture

7495 Systems WD2.0 of ISO/IEC 23004-2 Multimedia API

7496 Systems WD2.0 of ISO/IEC 23004-3 Component Model

7497 Systems WD2.0 of ISO/IEC 23004-4 Resource and Quality Management

7498 Systems WD2.0 of ISO/IEC 23004-5 Component Download

7499 Systems WD2.0 of ISO/IEC 23004-6 Fault Management

7500 Systems WD2.0 of ISO/IEC 23004-7 System Integrity Management

7501 Systems Extension and Maintenance Methodology of MPEG Standards

7502 Systems White Paper on MPEG-1 and MPEG-2 Systems

7503 Systems White Paper on MPEG-2 IPMP

7504 Systems White Paper on MPEG-4 Systems

7505 Systems White Paper on MPEG-4 IPMP

7506 Systems White Paper on MPEG-4 BIFS

7507 Systems White Paper on LASeR

7508 Systems White Paper on Font Compression and Streaming

7509 Systems White Paper on MPEG-7 Systems

7510 Systems White Paper on Multimedia Middleware

7511 Convener Ad Hoc Group on XML Structure Binarization and Streaming

7512 Convener Ad Hoc Group on Scene Representation

7513 Convener Ad Hoc Group on MPEG File Formats

46

7514 Convener Ad Hoc Group on MPEG Multimedia Middleware

7515 Systems White Paper on Streaming Text

7516 Video Request for subdivision of ISO/IEC 23002: Part 1 Accuracy Requirements forImplementation of Integer-Output 8x8 Inverse Discrete Cosine Transform

7517 Systems Request for minor enhancement 14496-1/Amd.2

7518 Audio New Work item Proposal on MPEG Audio Technologies

7519 Systems White Paper on Open Font Format

Annex 5Requirements

Source: Fernando Pereira (Instituto Superior Técnico, Lisboa-Portugal)

1 MPEG GENERAL

1.1 MPEG Dictionary12209, Mike Rubinfeld, MPEG VocabularyThis contribution proposes the creation of a dictionary and list of acronyms, in various phases, addressing terms inside and outside MPEG. The consensus has been to address for the moment only MPEG normative terms and targeting with the MPEG-21 standard. Other MPEG standards will follow if success is reached for MPEG-21. Following this agreement, an AHG on MPEG Dictionary chaired by Mike Rubinfeld (miker@nist.gov) has been created with the tasks to 1. Determine a set of normative terms & acronyms for MPEG-212. Collect terms with their associated definitions3. Point out inconsistencies in terminology used across MPEG-21 Standards4. Define the best clustering for the terms

2 MPEG-4

2.1 MPEG-4 SNHC Profiling (joint with the SNHC subgroup)12291 Jeong-Hwan Ahn, Mahnjin Han, Seong Won Ryu, on behalf of Mobile 3D Standardization Forum in Korea, Conformance Test of Simple 3D Compression Profiles and LevelsFollowing the decisions from previous meetings, this meeting continued to address the creation of a profile in the new 3D Compression profiling dimension. The 3D Compression Object Types, Profiles and Levels can be used in a stand alone stream or fitted in the MPEG-4 Architecture. The updates to the current Simple 3D Compression in the MPEG-4 Profiles under Consideration document proposed in m12291 have been accepted with a few changes. The final name adopted for the profile is Core. This Core 3D Compression and 2 levels have been promoted to an amendment of MPEG-4 AFX and thus will be removed from the next MPEG-4 Profiles under Consideration document.

47

2.2 MPEG-4 Laser Profiling (joint with Systems)12382, Jean-Claude Dufourd, Updated LASeR mini and full profile definitions

2.3 After accepting the small refinements proposed to Mini and Full Laser profiles (in the Scene Graph profiling dimension) included in the MPEG-4 Profiles under Consideration from last meeting, these 2 profiles, each with one level, have been promoted to an amendment of MPEG-4 Part 20 (Lightweight Application Scene Representation). Following this decision, the MPEG-4 Profiles under Consideration document will be updated by removing the Laser profiles under consideration).

2.4 MPEG-4 Audio Profiling (joint with Audio)12372, Ralph Sperschneider, Manfred Lutzky, Proposed new profile for high-quality communicationThis contribution proposed a new MPEG-4 Audio profile called ‘AAC Low Delay’ and a single level top address bidirectional communications using AAC with low delay. The profile uses a single object type, already defined. Since there was a clear industry needs and a good support was demonstrated by the industry, the profile was promoted to an amendment of MPEG-4 Audio.At the same meeting, it was agreed to update the MPEG-4 Profiles under Consideration document in the Audio area by removing the proposed profiles for which no work happen for some time.

2.5 Open Font Format Representation (joint with Systems)12113, Mike Ksar, Vladimir Levantovsky, AHG on Font Format Representation12169, Mike Ksar, Updated Working Draft of Open Font Format SpecificationThis problem has been solved with the submission by Microsoft and Adobe of a document addressing the request regarding disclaiming the copyright on the Open Font Format specification. This specification has been promoted to CD of a new part of MPEG-4 (Part 22). See the Systems report for more details on this issue.

3 MPEG-7

3.1 Content-based Search Framework (joint with Video)12114, Wo Chang, Tanya Beech, AHG on Content-Based Search Framework12206, Wo Chang, White paper on CBSearch FrameworkFollowing the discussions at the last meeting and the creation of an AHG in this area, the group tried again to understand what precisely the proposal implies, especially in normative terms. It was also considered that this task may only go forward if more people show interest since currently there is very little interest. In a last attempt to better identify this proposal, it was decided to create an HAG chaired by Wo Chang, Riccardo Leonardi, and Matthias Gruhne 1. To understand the scope of the CBSearch framework notablya. Objectivesb. Application scenariosc. Requirementsd. What needs to be normativee. Market needs and industry support2. To collaborate between WG1 and WG11 on the content-based search framework

48

4 MPEG-21

MPEG-21 General (joint with MDS and Systems)12180, Peder Drege, Thomas Skjølberg, Andrew Perkis, MPEG-21 client and server solutions for rich mobile content deliveryThis contribution describes a standards- based approach to the problem of delivering content to a diverse, and infinite, variety of mobile terminals. The solution developed is based on the MPEG-21 multimedia framework, with a special focus on the parts giving rich and standardized descriptors enabling media adaptation. Part 7 of MPEG-21, Digital Item Adaptation, allows the content provider to control the capabilities of the terminal receiving the content. The relevance for the DI Streaming work of this demo was recognized and the showed interest in starting following and contributing to these activities.

DI Streaming Requirements (joint with MDS and Systems)After the evaluation on the Sunday before the MPEG meeting of the answers to the Call for Proposals on DI Streaming issued at the last meeting, the work along the week concentrated on:

1. Processing of the outcome of the evaluation of the answers to the DI Streaming Call for Proposals – The outcome of the evaluation process was marked by the decision to not select any technology as straw man to address any requirement. This happened mainly because of the following reasons:

a. Some requirements need clarification before technology to address them is selectedb. For some other requirements, there were competing technologies that need to be

compared through core experiments before a decision is takenc. Still for other requirements, there is the need to compare the proposed technologies

with technologies already specified in MPEG standards 2. Improvement and clarification of the requirements – Following the conclusions of the

evaluation, improvements and clarifications to many requirements have been approved leading to a new version of the document ‘Requirements, Terminology and Use Cases for Digital Item Streaming’ (7278); still other requirements will be clarified and improved by the AHG in DI Streaming.

3. Analysis of usage cases versus requirements and tools received as feedback to the Call for Proposals – This analysis allowed to understand how the needs of the various identified usage cases are covered by the technologies under discussion.

4. Definition of Core Experiments for the cases where alternative tools exist – For the requirements addressed by competing technologies, core experiments have been defined. These core experiments are under the responsibility of the MDS subgroup.

Requirements on XML Fragment Requests (joint with MDS and Systems)12404, Stephen Davis, Ian Burnett, Joseph Thomas-Kerr, FRUs - Suggested Text for Fragment Request UnitsThis contribution proposes tools to address the draft requirements on XML fragment requests identified at last meeting (N7067). The requirements were reviewed and since no changes have been proposed, a document with final version of these requirements has been issued (N7279). To progress the work in this are, an AHG under the leadership of Joseph Thomas-Kerr was created to

1. Identify the appropriate location within MPEG standards for the XML Fragment Requests tools

2. Investigate whether existing technologies can meet the requirements established for XML Fragment Requests.

49

3. Validate that the proposed technology (FRU) meet the requirements established for XML Fragment Requests under conditions to be defined by the AHG and compare it with existing technologies.

REL Profiling (joint with MDS)12111, Xin Wang, Jaime Delgado, Chris Barlas, AHG on MPEG-21 REL ProfilingThe work developed in the AHG has been reported. Remind that at the starting of this meeting there was a profile under consideration called ‘REL Mobile’ profile.

12229, Jaime Delgado, Jose Prados, Eva Rodriguez, An MPEG-21 REL mobile profile12297, Jaime Delgado, Jose Prados, Eva Rodriguez, Comparison of REL mobile profiles These contributions propose a REL profile targeting the definition of a MPEG-21 REL profile matching OMA DRM v.2 specification, targeting easy conversion between the two REL specifications. The second contribution compares the proposed profile and the REL Mobile profile under consideration and concludes that it should be possible to merge the two profiles in a single one (what happened along the week).

12250, Satoshi Ito, Toru Kambayashi, Hideyuki Aisu, Hiroshi Niina, Usecase scenarios and REL Profile requirements for pre-recorded optical mediaThis contribution described the uses cases and proposes a REL profile for the area of pre-recorded optical media. It was considered that in principle this proposal should be able to be merged with the REL Mobile profile without much problems (what happened along the week).

12366, Xin Wang, Thomas DeMartini, Mai Nguyen, Manuel Ham, An Improvement of the MPEG REL Profile under Consideration in Response to the Requirements for Pre-recorded Optical MediaThis contribution shows how the profile proposed in the previous contribution could be merged with the REL Mobile profile under consideration (what happened along the week). The work targeting the creation of a single profile addressing as many application domains as posisble, led to the definition at the end of the week of the first REL profile (in fact first MPEG-21 profile) with the name of ‘Base REL Profile’ addressing a large set of applications, notably in the mobile and DVD domains. This profile will be defined through an amendment for the REL specification.

12213, Weongeun Oh, Yeonjeong Jeong, Taehyun Kim, Hogab Kang, MPEG-21 REL Profile for Broadcast This contribution proposes a REL profile for the broadcasting area. This proposal has been backed up by DMP through a liaison letter. Due to the effort to converge in the Base REL profile, there was much time left for the study of this profile but this will be a top priority at next meeting. However there is a clear need for more support to this profile and DMP was kindly asked to suggest its members to provide/show this support.

RDD Improvement (joint with MDS)Following the presentation of the technical contributions in the AHG meeting on the Sunday before the meeting, and some further discussions, the group agreed on the possibility to manage in order some improvements and clarifications are made to the RDD specification, very likely through an amendment (to be decided). These improvements should allow people not involved in the development of the standard to understand it better and address possible term inconsistencies/circularities, clarifications, etc. The more technical activities to be developed to reach this target are under the MDS subgroup responsibility.

50

5 MPEG-A

12109, Klaus Diepold, Wo Chang, AHG on MAFsThis contribution reported the activities developed in the context of AHG on MAF development since the last MPEG meeting.

12353, Herbert Thoma, Stefan Krägeloh, Harald Fuchs, Application Scenarios and Derived Requirements for Video Player MAF

This contribution proposes a Video Player MAF in some way parallel of the Protected Music Player under discussion. Although this proposal is still rather vague, it has been accepted for inclusion in the MAFs under Consideration document. Due to the similarities between this proposed MAF and the Super-distribution MAF already included in the MAFs under Consideration document, it is expected that either these two MAFs become rather different or their merging occurs.

12396, Jeongyeon Lim, Munchurl Kim, Synchronization of Multiple JPEG data to MP3 tracks in Music MAF Player Format

This contribution was presented to Requirements subgroup very late (Friday morning). It was agreed that this proposed MAF describes relevant application scenarios and it has a close relation with the Music Player MAF. The proposal of several related MAFs brings the important issue of the division of the MAF space. A major factor is this process is the relevance of the various proposals measured by the industry support brought for each of them.Regarding this proposal, it was agree that the AHG shall propose text for the description of the “Synchronization of multiple JPEG data to MP3 tracks in Music Player MAF format” for possible integration in the MAFs under Consideration document at the next meeting.

5.1 Protected Music Player MAF (joint with MDS, Systems, Audio)12360, Stefan Kraegeloh, Harald Fuchs, Proposed Protected Music Player MAF12197, Hendry, Munchurl Kim, Protecting and Governing Music MAF Player Format based Contents by using MPEG-21 IPMP12289, Zhongyang Huang, Shengmei Shen, Takafumi Ueno, IPMP Protection for MPEG-A Music Player12310, Jongheum Kim, Sungphil Heo, Weongeun Oh, Haekwang Kim, IPMP metadata protection for super-distribution MAF

These contributions address all the so called Protected Music Player MAF. After discussion, the authors of these contributions were able to make a common proposal for the revised version of the Protected Music Player MAF this means a version that would substitute the version of the Protected Music Player MAF issued at the last meeting.After long discussions, it was not possible to reach a consensus in the group regarding this common proposal. As a result of this disagreement, especially related to the choice of a specific encryption solution (AES128), it was agreed to update with the text of this common proposal the description of the Protected Music Player MAF in the MAFs under Consideration document (N7280) while highlighting the parts for which no agreement was reached. The AHG on MAFs should discuss these disagreements until the next meeting.

12335, Zvi Lifshitz, Cyril Concolato, TIRAMISU consortium, Proposal for free-distribution MAFThis contribution proposes a revision to the so-called Super-distribution MAF already included in the MAFs under Consideration document from last meeting. Although it was agreed to update the description of this MAF, this was not done because the author of this document could not be reached to implement this decision. This may happen at next meeting if the author is available.

51

5.2 Photo Player MAF (joint with Video)Regarding the Photo Player MAF, its requirements in the MAFs under Consideration document have been revised.Following the support by industry shown (Mitsubishi, Streamezzo, NIST), the Photo Player MAF was considered to have fulfilled all the conditions set by the Requirements subgroup for its promotion to MAF under consideration which means that this MAF has now been selected for specification and the remaining task regards its precise technical specification.

5.3 Summary on MAFsThe global MAF situation can be summarized as:

1. MAFs Specified (Final)i) Music Player MAF

2. MAFs Under Developmenti) Photo Player MAF

3. MAFs Under Considerationi) Protected Music Player MAFii) Super-Distribution MAF iii) Open Release MAFiv) Private Home Server MAFv) Video Player MAF

6 Explorations

6.1 Multi-view Video Coding (joint with Video)12328, Kwanghoon Sohn, Yongtae Kim, Changseob Park, Jaeho Lee, Comments on requirement and camera parametersFollowing the contribution above, the Multiview Coding Requirements document has been revised and a new version released (N7282). This revised document will serve as the basis for the Call for Proposals issued at this meeting.

6.2 New Colour Spaces in MPEG Standards (joint with Video)12112, Teruhiko Suzuki, AHG on Colour Spaces12251, Teruhiko Suzuki, Requirements of new colour spaces for MPEG videoAlthough MPEG does not define colour spaces, this activity follows the request to consider in MPEG specifications the extended gamut colour space for consumer applications under definition by IEC TC100. This request is being addressed in a more comprehensive way studying in general which new colour spaces may be added to the relevant MPEG standards. For this, an AHG chaired by Teruhiko Suzuki has been set up (continuing the AHG from last meeting) with the tasks 1. To study new extended-gamut colour space in IEC TC100 (IEC 61966-2-4)2. To study requirements for new colour spaces in MPEG standards3. To investigate other relevant colour spaces4. To study solutions to support new colour spaces

52

5. To study further work plan on this activity6. To investigate the way to support new colour spaces in MPEG standardsIn order, the work already done and to be done in this activity is well known by everybody, a document entitled “Report of activities on supporting new colour spaces in MPEG standards” (N7283) which summarizes the background, motivation, objectives, candidates colour spacees, way to proceed, etc.

6.3 VCTRThe Requirements subgroup had a first meeting with the VCTR Video break out group to study the possibility to create a document with VCTR Requirements. In this meeting a plan to define requirements for the VCTR Framework has been defined and a draft initial structure for the VCTR Requirements Document has been agreed.

7 Various

7.1 New Video Coding Schemes12355, Marco Tagliasacchi, Stefano Tubaro, Joao Ascenso, Kannan Ramchandran, Rohit Puri, Distributed video coding: a functional analysisThis contribution proposed a functional analysis of distributed video coding which is a new video coding paradigm which has been gaining increased interest in recent years. The theoretical foundations of DSC date back to the 70’s, but only in the last few years these concepts have been turned into practical video coding schemes. The purpose of this document was informative and thus no further action was required.

Annex 6Systems

Source: Systems Chair and Break-out group ChairsTitle: Systems Meeting ReportStatus: DraftEditor: Olivier Avaro (Streamezzo)

Contributors: David Singer (Apple), Claude Seyrat (Expway), Alexandre Cotarmanac’h (France Telecom R&D), Jean-Claude Dufourd (Streamezzo).

1 OverviewThe main outputs of the meeting from the Systems Sub-group perspective are:

No. Title TBP AvailableX 13818-1 :2000/Cor.4 FlexMux Code Point7469 Text of ISO/IEC 13818-1:2000/COR4 No 05/07/29X 13818-1 :2000/Amd.6 Transport of Streaming Text over MPEG-27470 Request of Amendment of ISO/IEC 13818-1:2000 No 05/07/297471 Text of ISO/IEC 13818-1:2000/PDAM 6 No 05/08/15X 14496-1 :2004/MPEG-4 Systems7517 Request for minor enhancement 14496-1/Amd.2 No 05/07/29

53

7472 Text of ISO/IEC 14496-1:2004/PDAM2 3D Compression Profile Descriptor No 05/07/29X 14496-11 :2004/Cor.47473 Text of ISO/IEC 14496-11:2004/COR4 Node Coding Table (Extrinsic Camera

parameters)No 05/08/15

X 14496-12 ISO File Format7474 Text of ISO/IEC 14496-12/DCOR2 No 05/08/087475 WD of ISO/IEC 14496-12/Amd.2 (Description of timed metadata) No 05/08/08X 14496-15 AVC File Format7476 Text of ISO/IEC 14496-15/DCOR1 No 05/07/297477 VM1.0 of ISO/IEC 14496-15/Amd.2 (SVC File Format) No 05/08/15X 14496-17 Streaming Text 7478 DoC on ISO/IEC 14496-17/FCD No 05/07/297479 Text of ISO/IEC 14496-17/FDIS No 05/08/15X 14496-20 LASeR7480 Study Text of ISO/IEC 14496-20/FCD No 05/07/297481 TuC for ISO/IEC 14496-20 No 05/08/217482 Validation Experiment on LASeR Binary Encoding No 05/07/297483 Validation Experiment on LASeR streamSource No 05/07/29X 14496-22 Open Font Format Specification7484 Request for a new part of ISO/IEC 14496-22 No 05/07/297485 Text of ISO/IEC 14496-22/CD No 05/09/01X 15938-1 Systems7486 Study Text on ISO/IEC 15938-1/FPDAM2 (Fast Access Extension) No 05/07/297487 WD 1.0 on MPEG-7 Systems Amd.2 Conformance (Fast Access Extension) No 05/07/297488 MPEG-7 Systems reference software workplan No 05/07/297489 DoC on Text of ISO/IEC 15938-1/DCOR2 (Misc.) No 05/07/297490 Text of ISO/IEC 15938-1/COR2 No 05/07/29X 23001-1 BinXML7491 Study Text on ISO/IEC 23001-1/FCD No 05/07/29X 23004 MPEG MultiMedia Middleware7492 Request for a New Work Item Proposal on MPEG Multimedia Middleware No 05/07/297493 Evaluation report on submission to M3W Extended Call for Proposal No 05/07/297494 WD2.0 of ISO/IEC 23004-1 Architecture No 05/07/297495 WD2.0 of ISO/IEC 23004-2 Multimedia API No 05/07/297496 WD2.0 of ISO/IEC 23004-3 Component Model No 05/07/297497 WD2.0 of ISO/IEC 23004-4 Resource and Quality Management No 05/07/297498 WD2.0 of ISO/IEC 23004-5 Component Download No 05/07/297499 WD2.0 of ISO/IEC 23004-6 Fault Management No 05/07/297500 WD2.0 of ISO/IEC 23004-7 System Integrity Management No 05/07/29X Promotion ,Exploration and Maintenance7501 Extension and Maintenance Methodology of MPEG Standards No 05/09/057502 White Paper on MPEG-1 and MPEG-2 Systems No 05/07/297503 White Paper on MPEG-2 IPMP No 05/07/297504 White Paper on MPEG-4 Systems No 05/07/297505 White Paper on MPEG-4 IPMP No 05/07/297506 White Paper on MPEG-4 BIFS No 05/07/297507 White Paper on LASeR No 05/07/297508 White Paper on Font Compression and Streaming No 05/07/297509 White Paper on MPEG-7 Systems No 05/07/297510 White Paper on Multimedia Middleware No 05/07/297515 White Paper on Streaming Text No 05/08/157519 White Paper on Open Font Format No 05/07/29

54

2 General issues

2.1 GeneralThe meeting report from Busan has been approved.

2.2 List of standards under development

Pr Pt Edit. Project Description CfP WD CD FCD FDIS2 1 2000 Amd.6 Transport of Streaming Text 05/07 06/01 06/072 1 2005 3rd Ed. MPEG-2 Systems 05/07 05/104 1 200x Amd.2 3D Profile Descriptor

Extensions05/07 06/01 06/07

4 12 2005 Cor.2 Miscellanea 05/04 05/07 05/104 12 2005 Amd.2 Description of timed

metadata05/07 05/10 06/01 06/07

4 15 2004 Amd.1 Support for FREXT 04/10 05/01 05/04 05/104 15 2004 Cor.1 AVC File Format 05/04 05/07 05/104 15 2005 Amd.2 SCV File Format Extensions 05/07 05/10 06/03 06/074 20 2004 1st Ed. Lightweight Application

Scene04/03 04/07 04/10 05/10

4 22 2005 1st Ed. Open Font Format 05/07 06/01 06/077 1 2004 Amd.2 BiM Extension 04/10 05/01 05/04 05/10B 1 200x 1st Ed. Binary Format 05/01 05/04 05/10E 1 200x 1st Ed. Architecture 05/01 05/07 06/04 06/07 07/01E 2 200x 1st Ed. Multimedia API 05/01 05/07 06/04 06/07 07/01E 3 200x 1st Ed. Component Model 05/01 05/07 06/04 06/07 07/01E 4 200x 1st Ed. Ressource & Quality

Management05/01 05/07 06/04 06/07 07/01

E 5 200x 1st Ed. Component Download 05/01 05/07 06/04 06/07 07/01E 6 200x 1st Ed. Fault Management 05/01 05/07 06/04 06/07 07/01E 7 200x 1st Ed. System Integrity

Management05/01 05/07 06/04 06/07 07/01

2.3 Standing Documents

Pr Pt Documents No. Meeting2 1 White Paper on MPEG-1 and MPEG-2 Systems N7502 05/07 Poznan2 11 White Paper on MPEG-2 IPMP N7503 05/07 Poznan4 1 White Paper on MPEG-4 Systems N7504 05/07 Poznan4 11 White Paper on MPEG-4 BIFS N7506 05/07 Poznan4 13 White Paper on MPEG-4 IPMP N7505 05/07 Poznan4 13 MPEG IPMP Extensions Overview N6338 04/03 München4 17 White Paper on Streaming Text N7515 05/07 Poznan4 18 White Paper on Font Compression and Streaming N7508 05/07 Poznan4 20 Presentation Material on LASER N6969 05/01 Hong-

Kong4 20 White Paper on LASeR N7507 05/07 Poznan4 22 White Paper on Open Font Format N7519 05/07 Poznan

55

7 1 White Paper on MPEG-7 Systems N7509 05/07 PoznanE X MPEG Multimedia Middleware Context and

ObjectivesN6335 04/03 München

E X White paper on Multimedia Middleware N7510 05/07 Poznan

2.4 Mailing Lists Reminder

Topic Information Kindly Managed by

General Systems

List

Liste Reflector : gen-sys@lists.uni-klu.ac.atList-Subscribe:

http://lists.uni-klu.ac.at/mailman/listinfo/gen-sysmailto:gen-sys-request@lists.uni-klu.ac.at?subject=subscribe

List-Archive: http://lists.uni-klu.ac.at/pipermail/gen-sysList-Help: mailto:gen-sys-request@lists.uni-klu.ac.at?subject=help

University of Klagenfurt

BinXML

Liste Reflector : mpeg7-sys@lists.uni-klu.ac.atList-Subscribe:

http://lists.uni-klu.ac.at/mailman/listinfo/mpeg7-sysmailto:mpeg7-sys-request@lists.uni-klu.ac.at?subject=subscribe

List-Archive: http://lists.uni-klu.ac.at/pipermail/mpeg7-sysList-Help: mailto:mpeg7-sys-request@lists.uni-klu.ac.at?subject=help

University of Klagenfurt

File Format

Liste Reflector : mp4-sys@lists.uni-klu.ac.atList-Subscribe:

http://lists.uni-klu.ac.at/mailman/listinfo/mp4-sysmailto:mp4-sys-request@lists.uni-klu.ac.at?subject=subscribe

List-Archive: http://lists.uni-klu.ac.at/pipermail/mp4-sysList-Help: mailto:mp4-sys-request@lists.uni-klu.ac.at?subject=help

University of Klagenfurt

LASeR

Liste Reflector : mpeg-laser@lists.uni-klu.ac.atList-Subscribe:

http://lists.uni-klu.ac.at/mailman/listinfo/mpeg-lasermailto:mpeg-laser-request@lists.uni-klu.ac.at?subject=subscribe

List-Archive: http://lists.uni-klu.ac.at/pipermail/mpeg-laserList-Help: mailto:mpeg-laser-request@lists.uni-klu.ac.at?subject=help

University of Klagenfurt

2.5 DemonstrationsNone.

2.6 FAQThe FAQ were updated as needed.

56

2.7 AOBNone.

3 MPEG-2 Systems (13818-1)

3.1 13818-1:2000 Cor.4M12185: Summary of Voting on ISO/IEC 13818-1:2000/DCOR 4. No comment. Finalization of the corrigendum.

Technical Work Finalized

3.2 13818-1:2005 Amd.6 Transport of Streaming TextNone.

Technical Work in Progress.

3.3 13818-1:200x 3rd EditionNone.

Technical Work in Progress.

4 MPEG-4 Scene Description (14496-11)

4.1 14496-11:2005 Cor.4

4.1.1 Topics1. Node Coding Table

4.1.2 ContributionsM12186 : Summary of Voting on ISO/IEC 14496-11:200X/DCOR 4. No comment. Document finalized.

Technical Work Finalized

4.2 14496-11:2005 Amd.2

4.2.1 Topics1. 3D Profile Descriptor Descriptor

4.2.2 ContributionsM12293: Proposal of 3D Compression Profile and Level Indication. Used as input to produce updated text of the amendment.

Technical Work in Progress.

4.3 Exploration

M12224: Proposed Amendment to 14496-11 for SMR Integration and Support. G. Zoia (EPFL) and P. Bellini (DSI) presented contribution M12224 “Proposed Amendment to 14496-11 for SMR Integration and Support”. As SMR renders music media information in different ways, mainly visual, the original proposal is based on a node MusisScoreTexture that shares analogies with MovieTexture.

57

During discussion the issue of raster vs. vector graphics has been raised; being this a decoder dependent issue, Systems experts proposed to use another kind of nodes (a more general MusicScore node) to allow the two modes at the same time. The issue of a hierarchy of less complex nodes has also been raised.

Conclusion was that a new Part 11 amendment should be started to support this functionality inside MPEG-4. Following this discussion and offline further discussion with Systems experts during the week, the SMR task group decided to prepare a new contribution for the 74th meeting; this contribution will take into account current and possibly future remarks to have a more mature proposal to start a WD of a new Part 11 amendment at next meeting. A specific work item has been defined in the SMR AHG mandate.

Technical Work in Progress.

5 ISO File Format (14496-12)

5.1 14496-12/Cor.2

5.1.1 Topics1. Miscellanea

5.1.2 ContributionsM12233: Corrigendum issues for the ISO File format. ISO File Format Corr. and/or amendment (draft at last meeting, new items from Apple) We need a brand for a file structured where a top-level meta-atom is the only mandatory atom

(just like iso2/isom for a top-level moov atom). ‘mp21’ is too specific. The permission in Part 12 to have a top-level meta box is maybe under-documented? So if you have a moov atom you may also have a top-level meta, but if there is no moov the brand is unclear. Maybe we do need ‘iso3’? But generally yes, the brands should be documented in the sections that they belong to, and isom and iso2 should go with the moov atom section.

The ‘pasp’ box being optional (and if not present, says ‘look at the codec-specific declarations’) would be useful in at least some cases (e.g. mpeg-4 short headers, aka 263, where the codec doesn’t tell you either). If present, it is an over-ride (if the codec has a default etc.).

Document ‘clap’ as well (after the permission from video)…if missing, the clean aperture is 100%.

Technical Work in Progress.

5.2 14496-12/Amd.2

5.2.1 Topics1. Description of Timed Metadata

5.2.2 ContributionsM12332: Proposed amendments to 14496-12 and/or 21000-9 for storage of timed metadata. Used as the basis to produce amendment text WD/PDAM. There were many questions of detail but the concept is accepted. The ISO FF editor and proponent to prepare an integrated amendment text.

Technical Work in Progress.

58

6 MPEG-4 AVC File Format (14496-15)

6.1 14496-15:2004 Amd.1

6.1.1 Topics1. Support for FREXT

6.1.2 ContributionsNone. AVC File format amendment for FRext; the ballot is still open. => Editor to make contribution of the amendment text and incorporated text.

Technical Work in Progress.

6.2 14496-15:2004/Cor.1

6.2.1 Topics1. Miscellanea

6.2.2 ContributionsAVC File format corr.; with no new items proposed, decision to convert this from a draft into a DCOR.

Technical Work in Progress.

6.3 14496-15:2004/Amd.2

6.3.1 Topics1. SVC File Format Extensions

6.3.2 ContributionsM12345: Proposal for SVC File Format Specification. Sample, tracks, levelgroups, fgs/mgs, sampleentry… Two sample types. AVC sample (pure) and SVC sample (with SVC NALu). Two motivations: don’t send excess data to an AVC decoder, and maybe issues with pure SVC AUs (no AVC data) and an AVC decoder.

Tracks: the idea of sharing AVC NALus in the media data (mdat) is nice, except that the AVC track will have single-sample chunks and a large chunk offset table.

LevelGroups: grouping of NAL Units (not just samples) into layers. Recursively defined, new sample table box that gives the levels that one can extract. The description has absolute offsets for every NAL needed in every sample. Overlap with sub-sample, or can we use the NAL Units themselves?

(Principle: try to find some parts of the design general for any scalable coding, some for any scalable video, and some that is specific to this SVC).

We could really do with an agreed term for a ‘cuboid’ (we have ‘layer’ to identify a group of cuboids that builds out to a cube with for a defined top-corner quality/size/temporal).

‘Dead’ sub-streams are not needed by *any* higher sub-stream. (Actually it’s ‘barren’). In some cases, a sub-stream may not be needed for *some* higher sub-streams, however. A name for this case? (‘Comatose’?)

59

The sampleentry can either do full scalability (‘levels’) or grouped into layers.

Handling FGS. Split an FGS NALu into fragments, which helps you find the right points in a succession of frames that give a constant quality. But it kind-of converts FGS into MGS.

Maintaining constant quality in FGS is understood to be an open question for JVT.

Some of the design might be affected by e.g. an RTP design for SVC.

M12370: Proposed SVC file format working draft. Leverage AVC FF. Flexible number of tracks. Use sample groups to map NALus to layers. Scalability info only signaled when needed. Group cuboids into layers (operating points).

(Side question: trimming FGS on the cube; which FGS cuboids get trimmed?)

isFGSlayer seems attached to a group of cuboids, but is a property of a given cuboid. Also flags for ROIlayer and sub-picture layer? A sub-picture layer is using slice-groups at the AVC level.

Discusses the layer dependencies (but is that really cuboid dependency?) and also which parameter sets are needed for a given layer.

We could do with a diagram of the track structure and how the samples relate (‘sub access units’ etc.)

This proposal was further clarified later in the meeting.

M12378: Proposed WD for the SV File Format. Layers again. Each sub-sample is associated with a layer. New sample table box associates each NALu with a layer.

Has two possible base definitions, depending on whether there is an AVC base or not.

Has association with NALus but there may be many NALus per cuboid.

M12232: Considerations for the SVC File format. Seems to build on previous Siemens idea? Maybe we could use actual hint tracks but have an ‘svc’ protocol? But the proposal could be more space-efficient and also allows a ‘copy’ extraction of AVC.Could use switch (extractor) tracks for the layer switch question below?

Joint video/systems meetingPreparation of Q&A for the joint meeting:

1.a. Q: We could really do with an agreed term for a ‘cuboid’ (we have ‘layer’ to identify

a group of cuboids that builds out to a cube with for a defined top-corner quality/size/temporal). Is it ‘sub-stream’?

b. A: ‘XXX enhancement stream’ maybe (xxx=quality/temporal/spacial)

2.a. Q: ‘Dead’ sub-streams are not needed by *any* higher sub-stream. (Actually it’s

‘barren’). In some cases, a sub-stream may not be needed for *some* higher sub-streams, however. A name for this case? (‘Comatose’?)

b. A: how about discardable and ‘conditionally discardable’?

60

3.a. Q: Can we switch ‘layer’ at any time, or do we wait for an IDR, or are there SVC

SP? You can always switch down, but switching up? b. This is an active discussion. Maybe we could allow more ‘IDR’ pictures in

enhancement streams, or ‘switch’ pictures, or…

4.a. Q: Definition of access units? ‘All the data for one time instant’? b. A:

5.a. Q: How does an AVC decoder handle an AU with no AVC NALus in it? Is this

legal? It lacks a primary coded picture? b. Lots of discussion, but probably an open question still. The file format can’t

currently answer “at what times are there access units?” with the question “what kind of decoder are you?” We’ll fudge around this in the file format, but SVC probably needs to be really clear about what constitutes an AVC-compatible stream, and how. If change is needed, it may be easier (e.g.) to insert stuff in the stream that keeps an AVC decoder happy, and handle that in SVC, rather than amending AVC (e.g. an all-prediction picture in AVC that tagged to be ignored by SVC).

6.a. Q: Constant-quality trimming of an FGS stream. What’s the ‘cookbook’? b. A: There have been some techniques adopted for ‘MGS’ (e.g. indicating cur-points

by splitting FGS slices into multiple NALus) but maybe we need some helper information? An open question.

7.a. Q: What is the trimming of (a) intermediate FGS layers (i.e. there is a higher SNR

layer) -- we assume you take all the FGS before taking more SNR; or maybe the higher CGS layer doesn’t use the FGS and is more efficiently coded for that SNR; (b) if the lower temporal or spacial layer has an FGS layer with it, and the upper t/s layer also has FGS, which of the two FGS layers get trimmed? Both?

b. A: All temporal layers of the upper spacial layer if it’s not much thinning, but if you want more, it kinda depends…the fraction points might give some guidance? Open question again.

8.a. Q: Is there a minimum trim of an FGS NAL Unit? NAL header, slice header,

motion vectors, residuals? b. A: Until 10% is left, or the NAL Header is left, whichever is larger. (The 10% is a

rough estimate).

9.a. Q: What is the latest (actual, practical) schedule for CD? We expect at least a one-

meeting lag on the file format, is this acceptable? b. A: currently targeted at next meeting CD but a delay is likely; the lag sounds like a

good idea.

Quick review of the answers from the joint video meeting.

61

Tracks and Samples from Nokia, clarifying the treatment.

When compatible: each track contains all the base and enhancement streams starting from the bottom for each instant in time. Tracks may represent anything up to the full quality, but have removed discardable and conditionally discardable streams; so it may be true that there is no track with the full SVC stream.

The new proposal is that a stream might several samples with the enhancement pictures for that time in separate samples. This worries the group; it has zero-duration samples. It also might happen that the NALus in an AU might have different decoding times (at the SVC level) (which would contradict the definition of AU given above). Check again!

The proposal for zero-duration samples mean that the decoding info for any time is formed by concatenating them (or at least the ones you need); one assumes that we’d have to restrict the composition offsets to be identical. This implies an amendment to part 12.

Good point that if an enhancement stream isn’t needed, it is more efficient to skip the entire picture of that enhancement stream in each AU than to skip NALu-by-NALu.

We agreed to instruct the editors to form a VM within a week and for everyone in the group to help get it into shape as an output in two weeks. Here is what we agreed to put in (elements of all contributions in a true ‘form the best’ fashion!). There is no implication that this is a WD i.e. we accept criticism or change on any or all parts of it; there’s no ‘accepted technology’ yet, it’s just a document to refine.

Agreements for the VM: 1) If an SVC track is fully AVC compatible, then it may be labeled as AVC; otherwise SVC

(follow Sony’s lead). We do need to be clear about ‘fully AVC compatible’ means (further work needed).

2) We agree that though there may be multiple tracks, we require that there is at least one that contains the full SVC stream (including all discardable enhancements streams aka dead streams). It has the advantage that this represents the basic data, and any re-ordering or subsets can be built from it (i.e. the set of tracks could be reduced to this, and then different subsets re-analyzed and extracted from it and a new file built).

3) The form of the sample structure is as documented for AVC file format; length/NALu interleaved, with 1/2/4 byte lengths.

4) A sample for the full scalable track is necessarily the full AU as defined by SVC5) The form of the sample entry: as AVC or pure SVC or combination (two 4CCs) from Sony

proposal6) We use the intro material from Siemens (1-4, generally, with appropriate edits)7) We do a group description (section 4.3) based on the Nokia proposal, but adjusted to have a

mandatory box based on the Sony mandatory fields; we note that these groups associate with less than a whole AU, not whole samples as the current sample-to-group box does. This might be missing some of the flexibility from the Siemens proposal.

8) We agree a group can document one or more enhancement streams, or a part of one (though that last capability may depend on how we do the mapping)

9) We agree we’ll put in a group mapping for now based on the Sony proposal but with general offset/length from the Siemens LevelGroupDescription box. Make sure the design allows (as for sample groups) association with more than one group (of different types).

62

10) We agree that if only temporal scalability is used in a layer, the AVC avss/avll design could be used, but for combination scalability we need something new

11) If multiple tracks are used, they are members of an alternate_group, with some suitable selection technique between them.

12) The other tracks all represent sub-parts of the SVC scalability and are complete operating points. (Thinned with respect to NALus, param sets, and with appropriate config). They may be AVC-compatible (see above).

13) In the other tracks, extractor NALus are permitted (Apple proposal).14) Group structures may be used on any/all tracks

Open questions:1) for the association of ‘less than a whole AU’ to groups we have 2.5 proposals; either split

the sample into zero-duration samples and use the s-to-g box; or do sub-sample-to-group mapping using a NAL-specific structure (Sony) or sub-sample box (generalization of Nokia); we need a CE to explore the consequences, but we take one for now.

2) the proposal above has a separate group description entry for each group, with dependency information, so you have to recursively form the closure. Is this efficient? Is it better to have a tree?

3) extractor NALus currently must point to base data (no recursion) but we might get space gains etc. if they can recurse?

4) Maybe we could use actual hint tracks but have an ‘svc’ protocol instead of extractor tracks?

Technical Work in Progress.

7 Streaming Text (14496-17) 7.1.1 Topics

1. Coded Representation of Text Stream.

7.1.2 ContributionsM12171: Summary of Voting on ISO/IEC FCD 14496-17.2. All comments have been disposed of and the specification has been finalized.

Technical Work Finalized

8 LASeR (14496-20) 8.1.1 Topics

1. Lightweight Application Scene Representation

8.1.2 Contributions

A list of items still to be validated was reported by the LAseR AHG (See AHG report). Since some of these items were preconditions to promote the current study of FCD to FDIS, it is decided to delay promotion to FDIS in Nice.

M12148: The document is presented by ENST and has been noted.

M12173: LASeR encoding XML format. This document was presented and discussed initially at the AHG meeting in Paris. There are questions about the degree of formality that this format should

63

have: many think that this format should be standard, because it has consequences on using the generic BiM case. An action was taken to modify the validation schema to allow SVG validation with W3C means (official RNG or schema) and to reformulate the LASeR validation as an extension. The result of this action was submitted as part of M12174 and used as a basis for the schema used as output of Poznan. An action was taken by Expway to produce text based on this contribution for inclusion in the specification by next meeting.

M12174 : LASeR reference software: codec. Discussed, the schema included in this contribution will be uploaded to CVS with a description of the changes from the version of June 8th. The validation schema in this contribution will be checked for synchronization with the encoding schema and uploaded to CVS.

M12175: LASeR and SAF editor's study. This contribution was presented and discussed. It was used as the basis for the construction of the study N7480.

M12177: Updated input device emulation proposal. The document was presented and discussed at the AHG meeting incParis. Some slight modifications have been made to the proposed text. The group deemed the proposal appropriate for inclusion in v1 if NB comments and an implementation are provided.  A FNB comment asked for the inclusion of this technology in Poznan, and an implementation was provided as part of M12385. As a result, the cursor technology was promoted (back) to the study.

M12178: The contribution was reviewed. This contribution presents several issues on encoding. a) Issues on the BiM encoding

a. MaskEncoding is used in BiMb. In Custom Encoding the ListEncoding is used.

b) Streamezzo notes that further to the AhG they have changed the encoding. They think there has been a misunderstanding on the procedure for the encoding procedure.

c) Rare transformation a. Open questions : Does it presereve the structure of the document ?

d) Not all "same" (e.g. samePath…) constructs prove benefit. to be studied in the evaluation experiment

e) Intermediate language seems a problemf) Decoder is large that due to the reordering of attributes in lexicographic order

Efficiency of the decoder to be evaluated in evaluation experimentg) BC encoding has no provision for forward compatibility when adding new attributes to

old elements (solution could be anyttribute)h) BC encoding has no construct for private data.

M12178 reported a problem on BiM efficiency issues and proposed solutions to overcome them. M12178 solution could solve the code size problem (divides it by 2). This should be verified in evaluation experiment. As a consequence the assessment of Streamezzo is that compression problems seem to be fixed thanks to M12178. This will be further evaluated in the “Validation Experiment”.

M12181, M12411: Handling Video in LASER (contribution for the AhG June 27th and 28th) and update on movieSource

use of clipPath: we need to see text and validate it before accepting this part (revisited during the week)

events: discussed, very good progress, warning that having attributes means that you need to use a script for processing such event, answer that movieSource is designed to process these

64

events without script, the proposal is reformulated, there is a relationship with SVG events, beginEvent and endEvent should be recycled, fixed in SVG and clarified in LASeR, need to write down IDL and examples ASAP

Editorial: the name of events does not need to have LASeR in it because the event will be in the LASeR namespace. The definition of Session needs to be added, ection on Alexandre to provide a definition of session.

movieSource: discussed, a better explanation/description of the feature has been achieved, possibly source as multiple children, add playlist mode, audio: want to generalize the proposal, need some more text ASAP to rediscuss it.

Validation experiment has been set-up on this technology. Final decision will be taken in Nice. M12281: LASeR: Improvement of compression performances of BiM-Compatible SimpleEncoding. The contribution was presented. Since the AHG meeting in Paris, Expway implemented 2 optimizations. The gain is 17% on non cartoons and 23% on cartoons, over June BiM-compatible SDL:

Same*: an XSL transformation is provided to translate from an XML without same* constructs to an XML using same* constructs.

Rare: XSL transformation was not provided since the translation is implemented in the provided code, but Expway could provide one.

The group had lots of questions, and since consensus could not be reached, it was decided to hold the validation experiment described in N7482 to answer all the questions.

M12340: Editorial and technical inputs for LASeR study of FCD. Studied and disposed of. More specifically : - refresh scene: solved- timing model: solved (see below)- extraction of fonts: solved by using the private data mechanism- encodings: in progress- movieSource: in progress, text to be provided and evaluated - enabled: todo

timing model: The basis of our discussion is the SVG timing model for the authoring form and the scene tree state (answers to uDOM calls). We have:

- LASeR AU time stamp (timeStampResolution from mediaHeader)- media AU time stamp (timeStampResolution from mediaHeader)- begin: scene time line- end: scene time line- dur: duration (relative) - repeatDur: duration (relative)- clipBegin [smil]: media time line, or normal play time- clipEnd [smil]: media time line, normal play time- delay [laser]: duration (relative)

Use cases: seekable, non-seekableDecisions:

- binary representation of begin/end: number of ticks, based on encoding resolution, using vluimsbf5, offset from CTS of current AU (DS, CC, AC, PS)

- representation of durations: number of ticks, based on encoding resolution, using vluimsbf5- binary representation of normal play time: different encoding resolution from the LASeR

one, vluimsbf5, statically defined in the spec (1000/s) - encoding resolution of time values: add attribute to LASeRHeader as proposed in M12340.

65

- RefreshScene is accepted as needed to synchronize the scene time of multiple clients not connected at the same time, in a non-seekable scenario.

o add RefreshSceneo add optional absolute time to NewScene.

M12382 : Updated LASeR mini and full profile definitions. Presented in a joint with Requirements, accepted for inclusion in the current document.

M12383: LASeR conformance streams. The contribution was presented. Some files have been reclassified to another directory. The group is requesting for them to be put on the CVS. The relation to profiles is that this is a raw repository, to be documented and sorted into profile suites.

M12384: French NB comment on LASeR. This NB comment was presented and discussed. The recommendation to promote technologies from the TuC to the study was accepted. The recommendation to promote the LASeR profiles from the PuC to the study was accepted. The recommendation to use the input study as a basis for the creation of the next specification was accepted.

M12385: The contribution was presented, and a demonstration on a Motorola E1000 phone was done for audio and video playback in LASeR/J2ME. The provided Java player and codec are in sync in terms of encoding. This satisfies the request of the meeting in Paris to see an implementation before the cursor was accepted into the study.

M12386: Deleted.

M12395: Comments and considerations on LASeR Study of FCD. This contribution was presented and discussed. It contains minor fixes to the input study. The changes were accepted.

Discussion on the Validation ExperimentA validation experiment has been set-up on the LASeR binarization. The validation experiment defines the appropriate schedule and success criteria for a proper decision by the group on LASeR binarization.

It is decided to hold an intermediate AhG meeting at Apple's premises end of September (22nd-23rd).

Technical Work in Progress.

9 Open Font Format Specification (14496-22) 9.1.1 Topics

1. Font Representation

9.1.2 ContributionsM12169: Updated Working Draft of Open Font Format Specification. Taken as a starting point to produce CD text.

The systems group acknowledges receipt of the IPR letter from Adobe, M12412, and the copyright release statement from Microsoft and Adobe, a copy of which is included below:

66

   ”© 2004 Microsoft Corporation and Adobe Systems Incorporated.  All rights reserved.

Microsoft and Adobe hereby grants you permission to copy and display the OFFS (the “Specification”), in any medium without fee or royalty.

THE SPECIFICATION IS PROVIDED "AS IS," AND THE AUTHORS MAKE NO REPRESENTATIONS OR WARRANTIES, EXPRESS OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NON-INFRINGEMENT, OR TITLE; THAT THE CONTENTS OF THE SPECIFICATION ARE SUITABLE FOR ANY PURPOSE; NOR THAT THE IMPLEMENTATION OF SUCH CONTENTS WILL NOT INFRINGE ANY THIRD PARTY PATENTS, COPYRIGHTS, TRADEMARKS OR OTHER RIGHTS.

 THE AUTHORS WILL NOT BE LIABLE FOR ANY DIRECT, INDIRECT, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF OR RELATING TO ANY USE OR DISTRIBUTION OF THE SPECIFICATION.

The name and trademarks of the Authors may NOT be used in any manner, including advertising or publicity pertaining to the Specification or its contents without specific, written prior permission. Title to copyright in the Specification will at all times remain with the Authors.

No other rights are granted by implication, estoppel or otherwise. “

Technical Work in Progress.

10 MPEG-7 Systems (15938-1)

10.1 15938-1/Amd.2

10.1.1 Topics1. BiM Extension

10.1.2 ContributionsM12192 - Editors Comments on ISO/IEC 15938-1/FPDAM2The contribution shows many editorial fixes. This document has been adopted for the new study of ISO/IEC 15938-1/FPDAM2.

M12352 - Comments on ISO/IEC 15938-1:2002/FPDAM 2The contribution has been accepted and the proposition has been be added to the new study of ISO/IEC 15938-1/FPDAM2.

M12193 - Proposal for 15938-1 Amd2 conformanceThe document has been reviewed and served as the basis of conformance WD for amendment 2.

M12258 - MPEG-21 binarisation status with BiM reference softwareThe document has been reviewed and its proposal partially accepted in the DCOR2 as the wildcard support fixes.

Technical Work in Progress.

67

10.2 15938-1/Cor.2

10.2.1 Topics1. BiM Correction

10.2.2 ContributionsM12188 - Summary of Voting on ISO/IEC 15938-1:2002/DCOR 2All comments have been reviewed.

Technical Work Finalized

11 MPEG-B

11.1 23000-1 Binary Format

11.1.1 Topics1. BiM

11.1.2 ContributionsM12287 - Editor's Study of ISO/IEC 23001-1 (BiM)The document has been reviewed and all changes (editorial) have been accepted. The terminology in MPEG-B Part 1 will become less MPEG-7 oriented.

Joint meeting with Requirements on Fragment referencesM12404: Fragment references were proposed to be added to the MPEG-B specification. However, due to architecture misalignment, the technology will be studied on the DI streaming activity.

Technical Work in Progress.

12 MPEG-E Multimedia Middleware (23004)

12.1 Multimedia Middleware

12.1.1 Topics1. MPEG Multimedia Middleware

12.1.2 Contributions

During its 73rd meeting (Poznan, Poland, July 25-29, 2005) MPEG has reviewed the responses it received on the Extended Call for Proposals on Multimedia Middleware (M3W) ISO/IEC JTC1/SC29/WG11 N7253 “Extended Call for Proposals on Multimedia Middleware (M3W)” as released at the April 2005 MPEG meeting in Busan, Korea.

The Extended Call for Proposals on Multimedia Middleware (M3W) specifically asked for submissions related to Part 5, the “Multimedia APIs” part of the requirements document as well as for Part 7 and more specific related to the security requirements in section 7.2.3) of the “In-Operation APIs” as those parts showed consistently lower coverage in the first case or have been omitted in the response to the initial Call for Proposals.

68

The review was carried out according to the requirements stated in the Requirements Document Version 2.0 (ISO/IEC JTC1/SC29/WG11 N6835 “MPEG Multimedia Middleware Requirements v.2.0” as released at the October 2004 MPEG meeting in Palma de Mallorca, Spain.

Following the evaluation of the responses in both the M3W AHG Meeting (Sunday July 24, 2005 in Poznan, Poland) as well as during the M3W BOG of the MPEG systems group it has been concluded that:

The submissions satisfy the request that was stated in the extended call for proposals; Provide sufficient coverage for the missing elements that resulted after the evaluation of the

submissions to the initial call for proposals.

M12196: The first contribution from the Information and Communications University of Korea is on the M3W Service Manager, Service, and Metadata. The contributions mostly cover the infrastructure aspect in building the middleware and have similarities each others; hence, there are many areas that the proposed technologies are overlap. After discussion seeking for harmonization, it was decided that the contributions from Philips will be used as the base technology and missing technologies are to be added from ICU and UoF contributions. One of the ICU’s contributions is regarding the Service Manager, Hierarchical composition of Service, and Metadata definition. From the 72nd MPEG meeting in Busan, we have worked in enhancing the technologies that we are responsible for. In the contribution, ICU reports the progress in enhancing the technology for Service Manager, Service, and the Metadata. This work has been presented and discussed during M3W AdHoc Group meeting that were held in Eindhoven, The Netherlands on May 23rd, 2005. Based on the evaluation of this and the other contributions it became clear that will be multiple IDL syntaxes. This is not a problem as long as there is a possibility for translation between these syntaxes. It will most likely be possible to translate the different syntaxes to each other and the automation of the necessary translations will be investigated (proposed mandate 1 for the M3W AHG). The service manager as such, inclusive of the metadata, discussion has been completed and the M12196 submission has been included in version 1.0 of ISO/IEC 23004 WD part 3.

M12260: The second contribution is from the Universal Home Application Programmers Interface Forum and proposes additional functionality for M3W as a follow-up on the fact that MPEG concluded during its Busan meeting, when reviewing the submissions to the initial Call for Proposals, that there are still some parts of the M3W requirements that have not been satisfied by responses and has issued an extended Call for Proposals for the missing functionality. These requirements are related to Part 5 “Multimedia APIs” and Section 7.2.3 “Security Requirements” of the M3W requirements version 2. The submission is related to part 5 “Multimedia APIs” of the requirements for M3W (as stated in the output document ISO/IEC JTC1/SC29/WG11 N6835 “MPEG Multimedia Middleware Requirements v.2.0” from the October 2004 MPEG meeting in Palma de Mallorca, Spain and proposes additional functionality as requested in the extended M3W CfP. The submission refers to PVR (Personal Video Recorder) functionality that is under development in the UHAPI consortium and as such is not yet complete. MPEG is requested however to consider this contribution for inclusion in the next version of the M3W WD (Oct 2005). This proposal has been accepted for consideration and will be discussed when the actual specification will be provided by the UHAPI forum. The M3W Multimedia API (part 2 of the proposed standard) will, in accordance with submissions M11847 (initial submission) and M12260 (response to the extended call for proposals), be based in part on the then most recent release of the Universal Home API as defined by the UHAPI forum. As mentioned in section 4.1 it will be considered to convert the UHAPI standard's IDL etc to the relevant M3W counterpart.

69

M12341: The third contribution is from the Broadcasting Media Research Group of ETRI and proposes IPMP Interfaces for M3W. This proposal is a response to the call for proposals for M3W – ISO/IEC JTC1/SC29/WG11 N7253, “Extended Call for Proposals on Multimedia Middleware (M3W)” [1] from April 2005 MPEG meeting in Busan. In this document, we address IPMP service interfaces, which correspond to a part of the requirements for “Multimedia API” of M3W. The IPMP submission (Proposal 3, M12341) was reviewed based on the new / modified requirements that are part of the IPMP submission. The specification is (partly) based on these new requirements and as such a study needs to be made whether the proposal not only caters for MPEG-2 and MPEG-4, but equally for MPEG-21 IPMP. If necessary the proposal should be adapted to also cater for MPEG-21 IPMP. Next to this it should be studied whether the IPMP Tools can and/or should be mapped to M3W Services (so that the download of tools as described in the IPMP submission can/should be unified with the M3W Component Download as described in Part 5 of the proposed standard) (action point for proposed M3W AHG).

M12259: Updated version of the M3W WD. Use as input document for the production of a new WD.

M12168: Multimedia Middleware (M3W) introduction. Used as input document to produce an output white paper on M3W.

An issue that is related to a contribution to the previous meeting (remote invocation form the University of Florence) needed to be reevaluated during the 73rd MPEG meeting due to the fact that the basic M3W architecture has been clarified. This led to a reevaluation of the positioning of the remote capabilities. This reevaluation could not be completed due to the fact that not all the relevant experts were available. A detailed discussion is planned for the mid-term M3W AHG meeting that will be planned between the Poznan and Nice meetings (action point for proposed M3W AHG).

An M3W AHG for the period leading to the 74th MPEG meeting in Nice, October 2005 has been established and will have a mid-term as well as a meeting on the Sunday before the Nice MPEG meeting.

Technical Work in Progress.

13 Joint Meetings and ExplorationJoint meeting with Requirements, MDS and Audio on MPEG-A : M12289, M12396, M12207, M12310, M12335, M12353, M12360, M12274: See Req. meeting report.Joint meeting with Requirements and MDS on DI Streaming : M12180: See Req. and MDS meeting reports.

70

14 Latest References and Publication Status

Pr Pt Standard No. Issue Status Doc. with Purpose ISO Award

2 1 ISO/IEC 13818-1/Amd.7 Published 2000/12 Done2 1 ISO/IEC 13818-1:2000 (MPEG-2 Systems 2nd Edition) 00/12 Published 2000/12 Proposed2 1 ISO/IEC 13818-1:2000/COR1 (FlexMux Descr.) N3844 01/01 Pisa Published 2002/03 N/A2 1 ISO/IEC 13818-1:2000/COR2 (FlexMuxTiming_

descriptor)N4404 01/12 Pattaya Published 2002/12 N/A

2 1 ISO/IEC 13818-1:2000/Amd.1 (Metadata on 2) & COR1 on Amd.1

N5867 03/07 Trondheim

Published 2003/12 Proposed

2 1 ISO/IEC 13818-1:2000/Amd.2 (Support for IPMP on 2) N5604 03/03 Pattaya Published 2004/03 N/A2 1 ISO/IEC 13818-1:2000/Amd.3 (AVC Carriage on MPEG-

2)N5771 03/07

TrondheimPublished XXXX Proposed

2 1 ISO/IEC 13818-1:2000/Amd.4 (Metadata Application CP)

N6847 04/10 Palma FDAM ITTF to be published N/A

2 1 ISO/IEC 13818-1:2000/Amd.5 (New Audio P&L Sig.) N6585 04/07 Redmond

FDAM ITTF to be published N/A

2 1 ISO/IEC 13818-1:2000/COR3 (Correction for Field Picture)

N6845 04/10 Palma COR ITTF to be published N/A

2 1 ISO/IEC 13818-1:2000/COR4 (M4MUX Code Point) N7469 05/07 Poznan COR ITTF to be published N/A2 11 ISO/IEC 13818-1:2003 (IPMP on 2) N5607 03/03 Pattaya Published 2003/12 Proposed4 1 ISO/IEC 14496-1 (MPEG-4 Systems 1st Ed.) N2501 98/10 Atl. City Published 1999/12 Done4 1 ISO/IEC 14496-1/Amd.1 (MP4, MPEG-J) N3054 99/12 Hawaii Published 2001/11 Done4 1 ISO/IEC 14496-1/Cor.1 N3278 00/03

Noordwijk.Published 2001/11 N/A

4 1 ISO/IEC 14496-1:2001 (MPEG-4 Systems 2nd Ed.) N3850 01/01 Pisa Published 2001/11 N/A4 1 ISO/IEC 14496-1:2001/Amd.1 (Flextime) Published 2002/10 Done4 1 ISO/IEC 14496-1:2001/Cor.1 N4264 01/07 Sydney COR ITTF N/A

71

4 1 ISO/IEC 14496-1:2001/Cor.2 N5275 02/10 Shangai COR ITTF N/A4 1 ISO/IEC 14496-1:2001/Cor.3 N6587 04/07

RedmondCOR ITTF N/A

4 1 ISO/IEC 14496-1:2001/Amd.2 (Textual Format) N4698 02/03 Jeju Island

AMD ITTF N/A

4 1 ISO/IEC 14496-1:2001/Amd.3 (IPMP Extensions) N5282 02/10 Shanghai

Published 2004-05 N/A

4 1 ISO/IEC 14496-1:2001/Amd.4 (SL Extension) N5471 02/12 Awaji Published 2003/12 N/A4 1 ISO/IEC 14496-1:2001/Amd.7 (AVC on 4) N5976 03/10

BrisbannePublished 2004-08 N/A

4 1 ISO/IEC 14496-1:2001/Amd.8 (ObjectType Code Points) N6202 03/12 Hawaii AMD ITTF to be published N/A4 1 ISO/IEC 14496-1:200x/Amd.1 (Text Profile Descriptors) N7229 05/04 Busan PDAM ITTF Final Text

EditingN/A

4 1 ISO/IEC 14496-1:200x/Cor4 (Node Coding Table) N7473 05/07 Poznan PDAM ITTF to be published N/A4 1 ISO/IEC 14496-1 (MPEG-4 Systems 3rd Ed.) N5277 02/10

ShanghaiIS ITTF to be published Proposed

4 6 ISO/IEC 14496-6:2000 Published 2000/12 N/A4 8 ISO/IEC 14496-8 (MPEG-4 on IP Framework) N4712 02/03 Jeju Published 2004-05 Proposed4 11 ISO/IEC 14496-11 (MPEG-4 Scene Description) N5279 02/10

ShanghaiFDIS SC29 Final Text

EditingProposed

4 11 ISO/IEC 14496-11/Amd.1 (AFX) N5480 02/12 Awaji FDAM ITTF Integration in 1st

Ed.N/A

4 11 ISO/IEC 14496-11/Amd.2 (Advanced Text and Graphics) N6205 03/12 Hawaii FDAM ITTF Integration in 1st

Ed.N/A

4 11 ISO/IEC 14496-11/Cor.1 N6203 03/12 Hawaii COR SC29 N/A4 11 ISO/IEC 14496-11/Cor.3 Valuator/AFX related correction N6594 04/07

RedmondCOR ITTF Integration in 1st

Ed.N/A

4 11 ISO/IEC 14496-11/Amd.3 Audio BIFS Extensions N6591 04/07 Redmond

FDAM ITTF Integration in 1st

Ed.Proposed

4 11 ISO/IEC 14496-11/Amd.4 XMT and MPEG-J Extensions N6959 05/01 HongKong

FDAM ITTF Integration in 1st

Ed.N/A

4 11 ISO/IEC 14496-11/Cor.3 (Audio BIFS Integrated in 3rd N7230 05/04 Busan COR ITTF Final Text N/A

72

Edition) Editing4 12 ISO/IEC 14496-12 (ISO Base Media File Format) N5295 02/10

ShanghaiPublished 2004-02 Proposed

4 12 ISO/IEC 14496-12/Amd.1 ISO FF Extension N6596 04/07 Redmond

FDAM ITTF FDAM 04/11/30 N/A

4 12 ISO/IEC 14496-12/Cor.1 (Correction on File Type Box)

N7232 05/04 Busan COR ITTF Final Text Editing

N/A

4 13 ISO/IEC 14496-13 (IPMP-X) N5284 02/10 Shanghai

IS ITTF to be published Proposed

4 14 ISO/IEC 14496-14 (MP4 File Format) N5298 02/10 Shanghai

Published 2003-11 Proposed

4 15 ISO/IEC 14496-15 (AVC File Format) N5780 03/07 Trondheim

Published 2004-04 Proposed

4 17 ISO/IEC 14496-17 (Streaming Text) N7479 05/07 Poznan FDAM ITTF TBP4 18 ISO/IEC 14496-18 (Font Compression and Streaming) N6215 03/12 Hawaii Published 2004-07 Proposed4 19 ISO/IEC 14496-19 (Synthesized Texture Stream) N6217 03/12 Hawaii Published 2004-07 Proposed7 1 ISO/IEC 15938-1 (MPEG-7 Systems) N4285 01/07 Sydney Published 2002/07 Done7 1 ISO/IEC 15938-1/Amd.1 (MPEG-7 Systems Extensions) N6326 04/03 Munich FDAM ITTF FDAM 04/11/28 N/A7 1 ISO/IEC 15938-1/Cor.1 (MPEG-7 Systems Corrigendum) N6328 04/03 Munich COR Editor N/A7 1 ISO/IEC 15938-1/Cor.2 (MPEG-7 Systems Corrigendum) N7490 05/07 Poznan COR ITTF N/A7 2 ISO/IEC 15938-2 (MPEG-7 DDL) N4288 01/07 Sydney Published 2002/02 Done21 9 ISO/IEC 21000-9 (MPEG-21 File Format) N6975 05/01

HongKongFDIS ITTF FDIS 05/01/21 Done

21 16 ISO/IEC 21000-16 (MPEG-21 Binary Format) N7247 05/04 Busan FDIS ITTF FDIS 05/04/22 TBP

73

15 Resolutions of Systems15.1.1 Cf. WG11 resolution.

16 List of Reviewed Contributions

N° Title Authors12149 Liaison Statement from 3GPP [SC 29 N 6742] 3GPP via SC 29 Secretariat12149 Liaison Statement from 3GPP [SC 29 N 6742] 3GPP via SC 29 Secretariat12168 Multimedia Middleware (M3W) introduction Jean H.A. Gelissen12169 Updated Working Draft of Open Font Format

SpecificationMike Ksar

12171 Summary of Voting on ISO/IEC FCD 14496-17.2 SC 29 Secretariat12171 Summary of Voting on ISO/IEC FCD 14496-17.2 SC 29 Secretariat12173 LASeR encoding XML format Jean-Claude Dufourd12174 LASeR reference software: codec Jean-Claude Dufourd12175 LASeR and SAF editor's study Jean-Claude Dufourd12177 Updated input device emulation proposal Jean-Claude Dufourd12178 Compression efficiency of the BiM-compatible SDL Jean-Claude Dufourd12180 MPEG-21 client and server solutions for rich mobile

content deliveryPeder Drege et al.

12181 Handling Video in LASER (contribution for the AhG June 27th and 28th)

Alexandre Cotarmanac'h et al.

12185 Summary of Voting on ISO/IEC 13818-1:2000/DCOR 4 SC 29 Secretariat12186 Summary of Voting on ISO/IEC 14496-11:200X/DCOR

4SC 29 Secretariat

12188 Summary of Voting on ISO/IEC 15938-1:2002/DCOR 2 SC 29 Secretariat12192 Editors Comments on ISO/IEC 15938-1/FPDAM2 Ray Taylor12193 Proposal for 15938-1 Amd2 conformance Ray Taylor12194 Bitstream Binding Language 2.0: A Response to the CfP

for Digital Item StreamingThomas-Kerr et al.

12195 Implementation System Documentation 2.0: Bitstream Binding Language

Thomas-Kerr et al.

12196 M3W Service Manager, Service and Metadata Hendry et al.12199 XML Streaming Instructions: response to the CfP for

Digital Item StreamingDevillers et al.

12200 Media Streaming Instructions: response to the CfP for Digital Item Streaming

Ransburg et al.

12201 Properties Style Sheet: response to the CfP for Digital Item Streaming

Devillers et al.

12203 Complex-mode Streaming Instructions: response to the CfP for Digital Item Streaming

Devillers et al.

12204 Signalling of buffer configuration parameters for DI Streaming: a response to the CfP on DI Streaming

Christian Timmerer et al.

12207 Propose Revised of WD 1.0 Photo Player MAF Wo Chang12210 Liaison Statement from OMA [SC 29 N 6794] OMA via SC 29 Secretariat12210 Liaison Statement from OMA [SC 29 N 6794] OMA via SC 29 Secretariat12224 Proposed Amendment to 14496-11 for SMR Integration

and SupportPierfrancesco Bellini et al.

12232 Considerations for the SVC File format David Singer12233 Corrigendum issues for the ISO File format David Singer

74

12258 MPEG-21 binarisation status with BiM reference software

Gregoire Pau et al.

12259 Updated version of the M3W WD Maaskant et al.12260 Reply to the M3W extended call for proposals Maaskant et al.12274 A MAF Photo Player for Enhanced Photo Browsing Sang-Kyun Kim et al.12281 LASeR: Improvement of compression performances of

BiM-Compatible SimpleEncodingPhilippe de Cuetos et al.

12287 Editor's Study of ISO/IEC 23001-1 (BiM) Philippe de Cuetos et al.12289 IPMP Protection for MPEG-A Music Player Zhongyang Huang e al. 12293 Proposal of 3D Compression Profile and Level Indication Jeong-Hwan Ahn et al.12310 IPMP metadata protection for super-distribution MAF Jongheum Kim et al.12332 Proposed amendments to 14496-12 and/or 21000-9 for

storage of timed metadataMichael Ransburg et al.

12335 Proposal for free-distribution MAF Zvi Lifshitz et al.12340 Editorial and technical inputs for LASeR study of FCD Cyril Concolato et al.12341 IPMP Interfaces of Multimedia API on M3W Hyon-Gon Choo et al.12345 Proposal for SVC File Format Specification Thomas Rathgen et al.12352 Comments on ISO/IEC 15938-1:2002/FPDAM 2 Andreas Hutter12353 Application Scenarios and Derived Requirements for

Video Player MAFHerbert Thoma et al.

12360 Proposed Protected Music Player MAF Stefan Kraegeloh et al.12363 Comments on DI streaming requirements Christian Timmerer et al.12370 Proposed SVC file format working draft Ye-Kui Wang et al.12378 Proposed WD for the SV File Format Mohammed Zubair et al.12382 Updated LASeR mini and full profile definitions Jean-Claude Dufourd12383 LASeR conformance streams Jean-Claude Dufourd12384 French NB comment on LASeR Jean-Claude Dufourd12385 LASeR utility software Jean-Claude Dufourd12386 Updated proposals for LASeR Technologies under

considerationJean-Claude Dufourd

12395 Comments and considerations on LASeR Study of FCD YeSun Joung et al.12396 Synchronization of Multiple JPEG data to MP3 tracks in

Music MAF Player FormatJeongyeon Lim et al.

12398 Liaison Statement from ECMA SC 29 Secretariat12398 Liaison Statement from ECMA SC 29 Secretariat12404 FRUs - Suggested Text for Fragment Request Units Stephen Davis et al.

75

Annex 7MDS

Source: Ian S Burnett, PhD, Chair

1.0 Overall Summary of MDS activity

MPEG-21 IPMP Components (Part 4): Discussions & MAF discussions New optional ‘protected’ content descriptor to be held in the clear – what is in the

‘protected package’ ? Output: MPEG-21 IPMP Components à Study of FCD

MPEG-21 RDD (Part 5): Discussions on terms, reviewing of inputs. Also Related Identifier changes for DII Output: Core Experiment on Abstraction and Resource, Related Identifiers

PDAM/1

2 pagers on DID, RDD, REL, DII, Schemas and MPEG-7 MDS….. All available out of this meeting

MPEG-21 Schemas output document Hosted on ITTF site Working Document – output kept up to date Update to DIP schema from this meeting

MPEG-21 DII Related Identifiers Amendment (Part 3): Inputs & Discussions

MPEG-21:REL (“Profiles”) - ReqtsRDD (Draft of AMD/1)DII (AMD/1)DIA AMD/1 (study of FPDAM)DIP (C++ Resources)Event Reporting (FCD)IPMP Components (Study of FCD)FragmentIDs (FCD)DI Streaming (WD)MAF DiscussionsMPEG-21 Schema Doc

76

Output: MPEG-21 DII Related Identifier Terms Amd 1 à Study of FPDAM, DoC

MPEG-21 Digital Item Adaptation (Part 7): Review CE results, Inputs, Comments BSDL and AQoS proposals to TuC – CEs out of Poznan Output: MPEG-21 DIA AMD/1 à Study of FPDAM, Working Draft of 2nd AMD

MPEG-21 Digital Item Processing (Part 10): Discussions of C++ Resources in DIs Output: PDAM of Amendment for C++ Bindings

MPEG-21 Event Reporting (Part 15): Review NB Comments ERR CE has resulted in some improvements in FCD Output: MPEG-21 Event Reporting à FCD, DoC

MPEG-21 Fragment Identification of MPEG Resources (Part 17) Review AHG results & Inputs Discussion on Fragment IDs for Streamed resources, IDs for AUs Output: MPEG-21 Fragment ID à FCD, DoC

MAF Discussions MDS input on IPMP for Protected Music Format IPMP Components experts involved in discussions Reqts report covered this – agreed output MAFs under consideration

MPEG-21 Digital Item Streaming Core Experiment on Fragmentation for DI Streaming CE on DI Mapping to Delivery channels

Includes work on use case from DMP Working Draft with terminology agreed Need Systems input

The MDS subgroup also organised a workshop on MPEG-21 on the Wednesday afternoon of the meeting. The workshop presentations are at: http://mpeg.nist.gov/mpeg/docs/73_Poznan/MtgReports/MPEG-21_Workshop.zip

The workshop agenda was:

14:00-14:20 Welcome - 5 years ago I had a vision Leonardo Chiariglione

14:20-14:40 Digital Items - the core concept and its declaration Ian Burnett

14:40-15:00 Digital Item Identification - how we know what ‘it’ is Niels Rump

15:00-15:20 IPMP Components Protected Digital Items Simon Watt

15:20-15:40 Rights Expression Language - Expressing Rights on usage of Digital Items Xin Wang 

BREAK - DEMOS

77

16:00-16:20 Rights Data Dictionary - Keeping track of all those termsChris Barlas

16:20-16:40 Digital Item Adaptation - Universal Multimedia Access in MPEG-21 Anthony Vetro

16:40-17:00 Digital Item Processing - Making Digital Items Smart Gerrard Drury

17:00-17:20 MPEG-21 Systems - File Format & DI Streaming Sylvain Devillers

DEMOS

2.0 Notes on discussions on Input Documents

The MDS subgroup commenced with a discussion regarding potential goals and intended outcomes for the week. The group then reviewed the outcomes of the AHGs.

12272 Myriam Amielh Sylvain Devillers Editor's input on ISO/IEC 21000-17 (FID)

12273 Myriam Amielh CE report on Logical Model Description

Input:The editors input contains only minor editorial changes and was not considered in detail as the text was reviewed later in the meeting. The CE expanded and considered in detail the logical model grammar functionality. Recommends adopting the logical model grammar and adopting several example models.

Actions:Discuss MP4 logical model with Systems. Adopt the logical model grammar proposed.

DI Streaming Joint Session with Requirements and Systems

12363 Christian Timmerer Michael Ransburg Hermann Hellwagner Comments on DI streaming requirements

12194 Thomas-Kerr (CRC-SITUoW) Ian Burnett (UoW enikos) Bitstream Binding Language 2.0: A Response to the CfP for Digital Item Streaming

12195 Thomas-Kerr (CRC-SITUoW) Ian Burnett (UoW enikos) Implementation System Documentation 2.0: Bitstream Binding Language

12199 Devillers Ransburg Timmerer Hellwagner Cotarmanac'h XML Streaming Instructions: response to the CfP for Digital Item Streaming

12200 Ransburg Devillers Timmerer Hellwagner Cotarmanac'h Media Streaming Instructions: response to the CfP for Digital Item Streaming

12201 Devillers Ransburg Timmerer Hellwagner Cotarmanac'h Properties Style Sheet: response to the CfP for Digital Item Streaming

12203 Devillers Ransburg Timmerer Hellwagner Cotarmanac'h Complex-mode Streaming Instructions: response to the CfP for Digital Item Streaming

12204 Christian Timmerer Hermann Hellwagner Andreas Hutter Jörg Heuer

Signalling of buffer configuration parameters for DI Streaming: a response to the CfP on DI Streaming

The session commenced with a summary of the report of the initial evaluation performed by the AHG over the weekend prior to the meeting.

Some requirements were noted as not being addressed for reasons of having been believed to be out of scope or being redundant. The decision was to take those requirements and reconsider them during the week with an aim of producing a revised set of requirements as a meeting output.

78

The two main proponents gave presentations clarifying their proposals. A BoG was set up to improve requirements, consider use cases, and draft Core Experiments dealing with non controversial requirements.

Tuesday

Joint Meeting with Requirements on REL Profiles

There are three domains with requirements. The key is to decide how many and how to nam ethe profiles for these domains. One approach is to have a common profile and an attribute on a license to indicate the ‘type’ of the REL profile. This gives just one profile with a ‘type’ which gives the sub-profile. However this approach is inconsistent with previous MPEG profiling. Another approach is that of multiple schemas. It would appear that the latter approach would result in simpler terminals. A further option is two profiles: Basic and Enhanced. The former is less complex and then the issue is the boundary between the two. The key is that whichever approach is chosen must be driven by the applications.

12229 Jaime Delgado Jose Prados Eva Rodriguez An MPEG-21 REL mobile profile

Input:The input is similar to the contribution made in Busian but in this case the content of the latter has been put into the format of the current thinking on REL profiles. It provides a comparison of REL profile options.

Actions:Work on issues during BoG.

12213 Weongeun Oh Yeonjeong Jeong Taehyun Kim Hogab Kang MPEG-21 REL Profile for Broadcast

Input:The input is related to a prior presentation to the Digital Media Project. It points out several perceived edeficiencies in the REL for broadcast purposes. It thus proposes several extensions and profiles of conditions/rights suitable for broadcasting.

Actions:One question was on how the profile will be applied. Currently this aims at the requirements of the DMP. The suggestion is that it is relevant to IP streaming applications. There was concern expressed that there had not been enough discussion with users/broadcasters. The proposal needs support from relevant users.

12250 Satoshi Ito Toru Kambayashi Hideyuki Aisu Hiroshi Niina Use case scenarios and REL Profile requirements for pre-recorded optical media

Input:The input discusses the ways in which REL could be used for pre-recorded optical media such as HD-DVD. It suggests several use cases and new reqts.

Actions:The actions were deferred until the related, next document was discussed.

79

12366 Xin Wang Thomas DeMartini Mai Nguyen Manuel Ham

An Improvement of the MPEG REL Profile under Consideration in Response to the Requirements for Pre-recorded Optical Media

Input:The view expressed in the input is that there are ways to meet the Optical disc requirements. The first five are general and some need to be in the REL profile and others in a ‘Title usage’ file.

Actions:It is suggested that the requirements are not dissimilar to other profile requirements. The BoG will consider the profiles during the week.

The BoG will consider the mobile profile differeneces, and the analyse the optical requirements and consider courses of action .

12299 Jaime Delgado Eva Rodriguez Maria Cirera Core Experiment on use of Event Report Requests: Specification of Use Cases

12356 Jaime Delgado Eva Rodriguez Maria Cirera Core Experiment on use of Event Report Requests: Software Implementation

Input:These two inputs on the CE on ERRs were presented in a single presentation. The CE investigated ERRs and implemented Software as two modules: Event watchdog and ER builder. Suggests that ERRs shouldn’t be mandatory to get an ER and that they can equally be generated by trusted devices without an ERR. Discussions required on how to adopt terms into ER.

Actions:Discussions regarding use of ERR in the FCD. The Spanish NB agreed to update their NB comment. The proposed changes will be implemented.

Reference software for ER is underway and will be combined with the software produced during the CE.

DIA Session

The DIA session commenced with a summary of the earlier DIA BoG (see slides below)

80

12217 Davy De Schrijver Wesley De Neve Rik Van de Walle Context-related attributes for MPEG-21 BSDL

Input:This document proposes context-related attributes for BSDL. The contribution considers the performance of the BintoBSD parser. BSDtoBin can be efficiently implemented but BintoBSD requires high memory footprint and high execution times. The proposal indicates significant improvements in speed and memory performance can be achieved using the proposed extensions. These were demonstrated using improvements to the reference software.

Actions:Proposal to create a CE on further investigations to see if this is the best way. One possibility is absolute Xpaths, restrictions on Xpaths etc. Also automatic computation of attributes was brought up as a possible issue.

81

12205 Sylvain Devillers Christian Timmerer Joseph Thomas-Kerr Report of CE on BSDL extensions

Input:The CE identified a set of new solutions for BSDL but has not selected final solutions. Some may require further work but others will likely achieve consensus during this meeting.

Actions:Discussions will take place during the meeting. The consensus solutions will be placed in the DIA TuC , those without consensus will be investigated further in CEs. Aim to get these into AMD/2 of DIA.

12222Paolo Nesi Davide Rogai Andrea Vallotti Pierfrancesco Bellini Jaime Delgado Eva Rodríguez

Report of CE on the indexing of IPMP protected DIDL content

Input:CE investigated an issue regarding ways to expose metadata held in protected Digital Items.Recommends introduction of an extra IPMP tag in IPMP DIDL.

Actions:There are questions re the usage of this approach for indexing. Also there are queries regarding the semantics of the descriptor for a hierarchy of protected content. A BoG was formed to consider the issues regarding the CE recommendations was formed to report back on Thursday morning.

Joint meeting on RDD with Reqts

12257 Marc Gauvin Jose Neri sDae on RDD Implementation Issues AHG

12172 Godfrey Rust Chris Barlas Niels Rump RDD Editors' comments on AhG Mandate Issues

12159 Godfrey Rust The Model of Making in indecs and RDD

12179 Leonardo Chiariglione M12159 revised and commented by lc

12358 Jaime Delgado Roberto Garcia Eva Rodriguez From Abstraction to Manifestation

12261 Melanie Dulong de Rosnay Contribution to AhG on MPEG-21 RDD

This meeting is recorded in the Requirements report in detail. The AHG discussed 2 documents in detail but did not discuss 12257, 12179, or 12261. 12179 is already incorporated and has been considered in discussions.

M12257 was presented.

The group decided to for a BoG to form a Corrigenda and Amendment. These at this meeting would be working drafts and would only contain consensual material. The process to be followed wil include CEs and reference software checkas and changes.

Joint Meeting on MAF with Reqts, and Audio

12310 Jongheum Kim Sungphil Heo Weongeun Oh Haekwang Kim IPMP metadata protection for super-distribution MAF

12335 Zvi Lifshitz Cyril Concolato TIRAMISU consortium Proposal for free-distribution MAF12360 Stefan Kraegeloh Harald Fuchs Proposed Protected Music Player MAF

12197 Hendry Munchurl Kim

Protecting and Governing Music MAF Player Format based Contents by using MPEG-21 IPMP

12289 Zhongyang Huang Shengmei Shen Takafumi Ueno IPMP Protection for MPEG-A Music Player

82

The outcomes of this meeting are recorded in the Audio and Requirements reports.

MDS continued meeting on DIA

12333Michael Ransburg Eric Delfosse Christian Timmerer Hermann Hellwagner

Proposed DIA extensions: AdaptationQoS for DI Streaming

Input:This input proposes an AQoS switch which enables fine grain resource description to allow dynamic adaptation on resource segments. This proposal needs to have support for single data element in the AQoS schema.

Actions:MDS will place the proposed schema change in the DIA TuC

12362 Michael Zufferey Eric Delfosse Christian Timmerer Harald Kosch

Proposed DIA extensions: Multiple input/output values for AdaptationQoS IOPins and MPEG-7 as BSDLink steering description

Input:This input proposes multiple I/O values for IO pins in Adaptation QoS. It proposes extensions to IOPIn sematics and syntax, and includes a choice between GetValue and GetVector elements. Proposes extended MPEG7SteeringDescriptionType.

Actions:MDS will place in the TuC and test the technology in a CE for the next meeting.

MDS – Wednesday

DII Discussions

12160 Niels Rump Editors input towards DoC on PDAM1 of 21000-3

12161 Niels Rump Editors input towards FPDAM1 of 21000-3

12162 Niels Rump for the UKNB Late UKNB comment on ISO/IEC 21000-3 PDAM1

12220 Niels Rump Chris Barlas Godfrey Rust Steffen Lindek Proposed additions to the RDD to satisfy DII/Amd.1

Input:The inputs relate to the ongoing DII proposals.

Actions:MDS will place 12161 as the study of PDAM. At the same time we will issue an RDD amendment to make explicit relationship types included in the RDD. IsA will be used for Annex D of the document. An editing period on the RDD amendment will fully populate the verbs. Amendment to be proposed and we request the Amendment. Issue a RDD PDAM and progress jointly DII and RDD PDAMs to FPDAM.

12218 Saar De Zutter Chris Poppe Frederik De Keukelaere Rik Van de Walle Contribution to WD of Amendment on C++ binding

12219 Frederik De Keukelaere Rik Van de Walle Contribution to DIP reference software - DIBO and DIXO implementations

Input:

83

The inputs look at DIP for executing C++ programs, using DIBOs from C++ programs, DIBO C++ bindings.

Actions:MDS created a BoG and the discussions were reported to the MDS group. Discussions are reported in the BoG report. One concern is platform independence and another is security. There was agreement that how C++ object code is application specific so need not be normative. Security applies whenever the .execute DIBO is called.

Amd containing C++ bindings for DIBOs, C++ GlobalEnv object and various extra statements regarding security issues when using DIP .execute. C++ executable resources can use DIBOs. Nothing normative for executing C++ resource executables from DIP.Intention is to produce a PDAM from this meeting.

DIP Conformance will be discussed in an AHG, as will profiling (perhaps a ‘safe’ profile). The reference software is looking good and should be complete within a few weeks of the current meeting. An intermediate version is on the ref software site and will be updated shortly.

The full BoG Report is given below:

Where: Room 14A/8When: Tue 1600-1730

Wed 1130-1200

17 Mandates

a) Discuss input contribution on C++ and make recommendations to MDS.

18 Recommendations

a) Begin a WD Ammendment 1 containing

1) C++ bindings for the DIBOs (including the object types)

2) C++ GlobalEnv object

3) C++ factory object to obtain DIBO implementations

4) Use of the C++ bindings, and invocation is application specific.

b) Include in the WD a "Security considerations" clause and also considerations for platform dependence regarding DIP.execute and use of C++ bindings. Include consideration of DIMs flagged as autoRun and calling DIP.execute.

c) Include example profile or update existing one to exclude DIP.execute and one to exclude C++ bindings.

Thursday :

Joint Session with Requirements

12180 Peder Drege Thomas Skjølberg Andrew Perkis MPEG-21 client and server solutions for rich mobile content delivery

84

Input:The document describes a commercial application which utilises MPEG-21 components – DID, DIA, DIP.

Actions:A discussion occurred regarding various issues brought out by the demonstration. Various suggestions on DIA extensions were suggested but it was agreed that we should take care on being too specific. The proponents intend participating in DI Streaming work and CEs.

12404 Stephen Davis Ian Burnett Joseph Thomas-Kerr FRUs - Suggested Text for Fragment Request Units

Input:Describes Fragment Request Units which allow fragments of DIs to be requested from a peer. The presenter noted the relationship to the previous presentation – mobiles requesting fragments of an XML document.

Actions:A discussion was held and it was decided that the best place for this work would be in MPEG-21 Part 18 DI Streaming.

DI Streaming Discussion

After a long discussion with Requirements on DI Streaming, it was decided to cluster the requirements in terms of use cases. Then for each use case to identify the way forward for each single tool required for the use cases – i.e. incoming proposed tool, multiple possibilities.

IPMP BoG report

Agreed a new ContentInfo ipmpdidl element which has vanilla semantics that could be used in given Application spaces. This will be added to the Study of FCD, there will be a Resolution requesting NBs to consider the latter during comments on the FCD and NB comments are required for the adoption of the text.

It was also noted that documents with editing periods require a banner as per last meetings UKNB position and resolution.

RDD BoG report

Dealt with an explicit issue – Abstraction/Manifestation. The group will create a CE to examine whether RDD supports requirements and/or modifications to RDD terms etc are required. It was

85

clearly expressed that the RDD structure will not be modified, but terms may be altered as long as all consequences and backward compatibility is considered.

REL Profiles BoG (with Requirements) report

Will generate a PDAM for MPEG REL with a three week editing period. The profile will be called the ‘base’ profile. Next meeting the broadcast profile proposal will be considered in more detail if there is an input. There are 5 companies currently supporting the Base profile.

MAF BoG (with Requirements and Audio) report

See the Audio Report for details on this discussion.

Notes on discussions

DI Streaming BoG (with Requirements) report

See the Requirements report for this discussion.

The Core experiments created by the DI Streaming BoG and the working draft are recorded in the resolutions (below) and the group also produced a simple Working Draft document.

3.0 MDS Resolutions

MDS Output Documents and Resolutions – Busan 73rd Meeting

MPEG-7

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available15938-5 MPEG-7 Multimedia Description Schemes

7418 Introduction to MPEG-7 MDS – an overview Yes 05/07/29

MPEG-21

The MDS subgroup notes that the latest set of 2 page MDS part introductions for the MPEG WWW site are held in documents N7418, N7420, N7422, N7423, N7427, N7433.

The MDS subgroup notes that the document N7419 is the second version of an ongoing working document containing the ‘electronic’ versions of schemas for the current MPEG-21 parts at IS/FDIS. The MDS subgroup requests that the electronic versions of the schemas be placed at a suitable (and publicised) location on the ITTF WWW site at the following URL: http://standards.iso.org/ittf/PubliclyAvailableStandards/MPEG-21_schema_files/

The MDS subgroup recommends approval of the following documents:

86

No. Title TBP AvailableMPEG-21 General

7419 Schema Files for MPEG-21 Standards Yes 05/07/297420 Introducing ISO/IEC 21000 Schema Files Yes 05/07/297421 MPEG-21 Workshop presentations Yes 05/07/29

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-2 MPEG-21 Digital Item Declaration

7422 Introducing MPEG-21 Digital Item Declaration Yes 05/08/19

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-3 MPEG-21 Digital Item Identification

7423 Introducing ISO/IEC 21000-3 DII Yes 05/07/297424 DoC of PDAM/1 on DII Related Identifiers No 05/07/297425 Study of PDAM/1 on DII Related Identifiers No 05/07/29

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-4 MPEG-21 IPMP Components

7426 Study of ISO/IEC 21000-4 IPMP Components FCD No 05/08/14

The MDS subgroup requests that NBs consider the Study of ISO/IEC 21000-4 IPMP Components FCD in making their NB comments on the FCD if at all possible.

The MDS subgroup reminds proponents of new technology included in the Study of ISO/IEC 21000-4 IPMP Components FCD of the requirement for NB comments supporting that technology at the next meeting in Nice.

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-5 MPEG-21 Rights Expression Language

7427 Introducing the MPEG REL Yes 05/08/057428 Request for Amendment 1 of ISO/IEC 21000-5 No 05/07/297429 PDAM/1 MPEG-21 REL Profiles No 05/08/19

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-6 MPEG-21 Rights Data Dictionary

7430 Request for Amendment 1 of ISO/IEC 21000-6 No 05/07/297431 PDAM/1 ISO/IEC21000-6 Digital Item Identifier Relationship Types No 05/08/317432 Workplan for CE on RDD Abstraction and Resource No 05/07/297433 Introducing the MPEG-21 RDD Yes 05/07/29

87

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-7 MPEG-21 Digital Item Adaptation

7434 Study of ISO/IEC 21000-7 FPDAM/1 No 05/07/297435 TuC for Potential 21000-7 MPEG-21 DIA Amendment(s) v3.0 No 05/08/057436 WD of ISO/IEC 21000-7 AMD/2 v1.0 No 05/08/057437 DIA Reference Software Status and Workplan No 05/07/297438 Workplan for CE on Adaptation QoS and BSDLink extensions No 05/07/297439 Workplan for CE on BSDL Extensions No 05/07/29

The MDS subgroup requests that NBs consider the Study of ISO/IEC 21000-7 FPDAM/1 in making their NB comments on the FPDAM if at all possible.

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-8 MPEG-21 Reference Software

7440 Introducing the MPEG-21 Reference Software Yes 05/07/29

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-10 MPEG-21 Digital Item Processing

7441 Request for Amendment 1 of ISO/IEC 21000-10 No 05/07/297442 PDAM/1 of ISO/IEC21000-10 DIP: Additional C++ Bindings No 05/08/29

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-15 MPEG-21 Event Reporting

7443 DoC on ISO/IEC 21000-15 CD MPEG-21 ER No 05/07/297444 FCD of ISO/IEC 21000-15 MPEG-21 ER No 05/08/29

The MDS subgroup thanks the National Bodies of Australia, Austria, Spain, Korea and USA for their NB comments on ISO/IEC 21000-15 CD.

The MDS subgroup remind proponents of technology included in the ISO/IEC 21000-15 FCD of the need for Reference Software to be submitted at the 74th MPEG meeting.

The MDS subgroup recommends approval of the following documents:

No. Title TBP Available21000-17 MPEG-21 Fragment Identification of MPEG Resources

7445 DoC on ISO/IEC 21000-17 CD MPEG-21 Fragment Identification of MPEG Resources

No 05/07/29

7446 ISO/IEC 21000-17 FCD MPEG-21 Fragment Identification of MPEG Resources

Yes 05/09/10

88

The MDS subgroup thanks the National Bodies of Australia, France, Japan, UK and USA for their NB comments on ISO/IEC 21000-17 CD.

The MDS subgroup remind proponents of technology included in the ISO/IEC 21000-15 FCD of the need for Reference Software to be submitted at the 74th MPEG meeting.

No. Title TBP Available21000-18 MPEG-21 Digital Item Streaming

7447 Working Draft of MPEG-21 DI Streaming7448 Workplan for CE on Fragmentation for DI Streaming 05/07/297449 Workplan for CE on DI Mapping to Delivery channels 05/07/29

The MDS subgroup thanks University of Klagenfurt, Siemens, France Telecom, DANAE, and the University of Wollongong for their submissions of technology in response to the DI Streaming call for technologies.

N7450 AHG on MPEG-21 DIPMandate: 1. Complete editing of text of DIP PDAM 1.

2. Carry out further study of DIP PDAM 13. Carry out study of DIP profiling.4. Carry out study of DIP conformance

Chairs: Gerrard Drury (gerrard at enikos.com)Frederik De Keukaleare (Frederik.DeKeukelaere at ugent.be)

Duration: Until the 74th MeetingMeetings: AhG Meeting to be held on the Sunday prior to 74rd meeting.

Other business to be conducted by email or phone.Reflector: mpeg-21-dip at uow.edu.auSubscribe: Follow procedure at http://mailinglists.uow.edu.au/mailman/listinfo/mpeg-21-dip

N7451 AHG on MPEG-21 DIAMandate: 1. Carry out Core Experiments on DIA and provide recommendations to

the MDS subgroup.2. Produce an editors' input on the Study of ISO/IEC 21000 FPDAM/13. Produce an editors' input on WD of AMD/2 v1.0.4. Maintain the MPEG-21 DIA TuC document.5. Maintain the DIA reference software status and work plan.

Chair: Christian Timmerer (christian.timmerer_at_itec.uni-klu.ac.at)Duration: Until the 74th meeting.Meetings:

AHG meeting will be held on the weekend prior to 74 th meeting. Other business will be conducted by e-mail or telephone conference.

Reflector: mpeg21-uma_at_merl.comSubscribe: To subscribe send email to avetro_at_merl.com (Anthony Vetro).

N7452 AhG on MPEG-21 Event Reporting

89

Mandate: 1. Determine the relation between Event Reporting and IPMP Digital Items2. Further develop Reference Software for Event Reporting3. Validate the relationship between Event Report Requests and their embedding

within DIsChairs: FX Nuttall (fx at nuttall.org)

Andrew Tokmakoff (tokmakoff at telin.nl) Duration: Until the 74th MeetingMeetings: A pre-meeting will be held on the weekend before the 74th meeting.

Other business will be conducted by E-mail or telephone conference.Reflector: mpeg-21-event-reporting@lists.telin.nl Subscribe: See http://lists.telin.nl/mailman/listinfo/mpeg-21-event-reporting

7453 AHG on MPEG-21 IPMP ComponentsMandate: 1. To review National Body input from the FCD ballot.

2. To make recommendations on the drafting of the FDIS to the 74th Meeting Chairs: Simon Watt (simon_dot_watt_at_umusic.com)Duration: Until the 74th MeetingMeetings: Meetings will be held (if required):

in Nice on the Sunday prior to the 74th MeetingReflector: mpeg-21@listserv.imec.be message titles prefixed [IPMP]Subscribe: To subscribe, go to http://listserv.imec.be/mailman/listinfo/mpeg-21

N7454 AHG on Digital Item StreamingMandates: 1. Refine DI Streaming requirements and use cases, in particular for IPMP

scenarios.2. Conduct Core Experiments on DI Streaming3. Further study the state of the art about buffering information and list

examples where such information is required4. Identify normative interfaces required for the mapping of streamed DIs to

delivery channels.5. Evaluate how existing technologies such as MPEG-2/4 Systems can be used

for the streaming of DIs6. Investigate the relevance of existing technologies for unfulfilled

requirements

Chairs: Sylvain Devillers (sdevillers.ext_at_rd.francetelecom.com)Joseph Thomas-Kerr (joetk_at_elec.uow.edu.au)Ingo Wolf (WolfI_at_t-systems.com)

Duration: Until 74th meetingMeetings: Week-end before the 74th meeting.Reflector: mpeg-21@xlistserv.imec.be messages to use the prefix [DIS] – remove x from

email addressSubscribe: Apply at http://listserv.imec.be/mailman/listinfo/mpeg-21 

90

4.0 MDS Poznan ScheduleMDS Sub-Group Schedule

  MPEG MDS Chair: Ian S Burnett    

  MPEG-7 and MPEG-21 v.4.0    

Number Source Title STD NOTES

         Monday Morning (9h00-13h00)        MPEG Plenary   Plenary room    

         Monday Afternoon (14h00-20h00)        

         Kick-off of MPEG MDS activities (14h00-14h30)   MDS Room 14A/8   ISB

 Agenda, Goals and Issues for the Week for MDS Group Ian S Burnett    

         Review of AHG resolutions, CE results and action points (14h30-15h45)   MDS Room 14A/8 ALL ISB

12133Gerrard Drury Frederik De Keukaleare AHG on MPEG-21 DIP    

12134 Niels Rump AHG on Functional Granularity of DII    

12135Christian Timmerer Thomas DeMartini AHG on MPEG-21 DIA    

12136 FX Nuttall Andrew Tokmakoff AHG on MPEG-21 Event Reporting    

12137 Myriam Amielh Sylvain Devillers AHG on MPEG 21 Fragment Identifiers    

12138 Marc Gauvin AHG on RDD Implementation Issues    

         Define BoGs and Mandates (15h45-16h00)   MDS Room 14A/8 ALL ISB  DIA, DIP, ER, FID Break Outs: 14A/18    

 

DIA - Study of FPDAM, Solns for BSDL, DIA AMD/2 tools, ref sw workplan Tuesday 9am-11am    

  DIP - Not yet scheduled      ER Tuesday pm, Wednesday am      RDD 15/208 Thursday 9:00-11:00    MPEG-21 FID (16h00 - 16h30)   MDS Room 14A/8 MPEG-21  

12272 Myriam Amielh Sylvain Devillers Editor's input on ISO/IEC 21000-17 (FID)    

12273 Myriam Amielh CE report on Logical Model Description             MPEG-21 DI Streaming (16h30 - 18h00)   REQTS 14B/3 MPEG-21  

12363Christian Timmerer Michael Ransburg Hermann Hellwagner

Comments on DI streaming requirements    

12194Thomas-Kerr (CRC-SITUoW) Ian Burnett (UoW enikos)

Bitstream Binding Language 2.0: A Response to the CfP for Digital Item Streaming    

12195Thomas-Kerr (CRC-SITUoW) Ian Burnett (UoW enikos)

Implementation System Documentation 2.0: Bitstream Binding Language    

12199Devillers Ransburg Timmerer Hellwagner Cotarmanac'h

XML Streaming Instructions: response to the CfP for Digital Item Streaming    

12200Ransburg Devillers Timmerer Hellwagner Cotarmanac'h

Media Streaming Instructions: response to the CfP for Digital Item Streaming    

12201Devillers Ransburg Timmerer Hellwagner Cotarmanac'h

Properties Style Sheet: response to the CfP for Digital Item Streaming    

12203Devillers Ransburg Timmerer Hellwagner Cotarmanac'h

Complex-mode Streaming Instructions: response to the CfP for Digital Item Streaming    

12204Christian Timmerer Hermann Hellwagner Andreas Hutter Jörg Heuer

Signalling of buffer configuration parameters for DI Streaming: a response to the CfP on DI Streaming    

         Tuesday Morning (9h00-13h00)                 REL Profile Issues with REQTS 14B/3 (9h00-11h00)   REQTS 14B/3 MPEG-21 ISB

12111 Xin Wang Jaime Delgado Chris Barlas AHG on MPEG-21 REL Profiling    

12213Weongeun Oh Yeonjeong Jeong Taehyun Kim Hogab Kang MPEG-21 REL Profile for Broadcast    

12229Jaime Delgado Jose Prados Eva Rodriguez An MPEG-21 REL mobile profile    

12250Satoshi Ito Toru Kambayashi Hideyuki Aisu Hiroshi Niina

Use case scenarios and REL Profile requirements for pre-recorded optical media    

12297Jaime Delgado Jose Prados Eva Rodriguez Comparison of REL mobile profiles    

92

12366Xin Wang Thomas DeMartini Mai Nguyen Manuel Ham

An Improvement of the MPEG REL Profile under Consideration in Response to the Requirements for Pre-recorded Optical Media    

         

MPEG-21 DID Comments (09h00 - 10h00)Event Reporting (11h00 - 12h00) MDS Room 14A/8 MPEG-21  

12299Jaime Delgado Eva Rodriguez Maria Cirera

Core Experiment on use of Event Report Requests: Specification of Use Cases    

12356Jaime Delgado Eva Rodriguez Maria Cirera

Core Experiment on use of Event Report Requests: Software Implementation    

         

MPEG-21 DID Comments (09h00 - 10h00)DIA Issues (12h00 - 13h00, 14h00 - 15h00) MDS Room 14A/8 MPEG-21  

12333Michael Ransburg Eric Delfosse Christian Timmerer Hermann Hellwagner

Proposed DIA extensions: AdaptationQoS for DI Streaming    

12191

Thomas DeMartini Christian Timmerer Benoît Pellan Truong Cong Thang Eric Delfosse

Report of CE on Harmonisation of Conversion Tools    

12205Sylvain Devillers Christian Timmerer Joseph Thomas-Kerr Report of CE on BSDL extensions    

12217Davy De Schrijver Wesley De Neve Rik Van de Walle

Context-related attributes for MPEG-21 BSDL    

12362Michael Zufferey Eric Delfosse Christian Timmerer Harald Kosch

Proposed DIA extensions: Multiple input/output values for AdaptationQoS IOPins and MPEG-7 as BSDLink steering description    

12364 Christian Timmerer Editors' input on Study of ISO/IEC 21000-7 FPDAM/1    

12158 SC 29 Secretariat Summary of Voting on ISO/IEC 21000-7:2004/FPDAM 1 [SC 29 N 6751]    

Tuesday Afternoon (14h00-20h00)                 

MPEG-21 IPMP (14h30 - 15h00)   MDS Room 14A/8 MPEG-21  

12222

Paolo Nesi Davide Rogai Andrea Vallotti Pierfrancesco Bellini Jaime Delgado Eva Rodríguez

Report of CE on the indexing of IPMP protected DIDL content    

         

MPEG-21 DIA AMD/1 Editing (13h00- 14h30)RDD Issues with REQTS 14B/3(15h00 - 16h00) REQTS 14B/3 MPEG-21  

12257 Marc Gauvin Jose Neri sDae on RDD Implementation Issues AHG    

12172 Godfrey Rust Chris Barlas Niels Rump RDD Editors' comments on AhG Mandate Issues    

12159 Godfrey Rust The Model of Making in indecs and    

93

RDD

12179 Leonardo Chiariglione M12159 revised and commented by lc    

12358Jaime Delgado Roberto Garcia Eva Rodriguez From Abstraction to Manifestation    

12261 Melanie Dulong de Rosnay Contribution to AhG on MPEG-21 RDD             MPEG-21 DIA AMD/1 Editing (13h00- 14h30) MAF (16h00 - 18h00) REQTS 14B/3 MPEG-21  

12310Jongheum Kim Sungphil Heo Weongeun Oh Haekwang Kim

IPMP metadata protection for super-distribution MAF    

12335Zvi Lifshitz Cyril Concolato TIRAMISU consortium Proposal for free-distribution MAF    

12353Herbert Thoma Stefan Krägeloh Harald Fuchs

Application Scenarios and Derived Requirements for Video Player MAF    

12360 Stefan Kraegeloh Harald Fuchs Proposed Protected Music Player MAF    

12197 Hendry Munchurl Kim

Protecting and Governing Music MAF Player Format based Contents by using MPEG-21 IPMP    

12289Zhongyang Huang Shengmei Shen Takafumi Ueno

IPMP Protection for MPEG-A Music Player    

12396 Jeongyeon Lim Munchurl Kim

Synchronization of Multiple JPEG data to MP3 tracks in Music MAF Player Format    

         MPEG-21 DID Comments (09h00 - 10h00) DIA Issues (18h15-19h00) MDS Room 14A/8 MPEG-21  

12333Michael Ransburg Eric Delfosse Christian Timmerer Hermann Hellwagner

Proposed DIA extensions: AdaptationQoS for DI Streaming    

12362Michael Zufferey Eric Delfosse Christian Timmerer Harald Kosch

Proposed DIA extensions: Multiple input/output values for AdaptationQoS IOPins and MPEG-7 as BSDLink steering description    

         Wednesday Morning (09h00-13h00)                 

MPEG Plenary (9h00-11h00)   Plenary roomMPEG Plenary  

         MPEG-21 DII (11h30 - 12h00)   MDS Room 14A/8 MPEG-21  

12160 Niels Rump Editors input towards DoC on PDAM1 of 21000-3    

12161 Niels Rump Editors input towards FPDAM1 of 21000-3    

12162 Niels Rump for the UKNB Late UKNB comment on ISO/IEC 21000-3 PDAM1    

12220 Niels Rump Chris Barlas Godfrey Rust Proposed additions to the RDD to    

94

Steffen Lindek satisfy DII/Amd.1

MPEG-21 DIP (12h00 - 12h30)   MDS Room 14A/8 MPEG-21  

12218Saar De Zutter Chris Poppe Frederik De Keukelaere Rik Van de Walle

Contribution to WD of Amendment on C++ binding    

12219Frederik De Keukelaere Rik Van de Walle

Contribution to DIP reference software - DIBO and DIXO implementations    

         Wednesday Afternoon (14h00-17h45)                 MPEG-21 Tutorial (14h00 - 17h45)   MDS Room 14A/8 MPEG-21  

                           MPEG Social Event (17h45 - midnight)               Thursday Morning (09h00-12h30)                                   MPEG-21 Systems/Applications (09h00 - 10h00)   REQTS 14B/3 MPEG-21  

12180Peder Drege Thomas Skjølberg Andrew Perkis

MPEG-21 client and server solutions for rich mobile content delivery    

12404Stephen Davis Ian Burnett Joseph Thomas-Kerr

FRUs - Suggested Text for Fragment Request Units    

         MPEG-21 DI Streaming Ces etc (10h00 - 12h00)   REQTS 14B/3 MPEG-21                             

MPEG-21 IPMP/DIP/RDD Issues (12h0 - 12h30) MDS Room 14A/8 MPEG-21  

         Thursday Afternoon (14h00-19h00)        MPEG-21 REL Profiles BoG report(14h00 - 15h00)   REQTS 14B/3 MPEG-21           MAF Protected Music Player - Update & approval (15h00 - 16h00)   REQTS 14B/3 MPEG-21  

95

         MPEG-21 DI Streaming Ces etc (16h00 - 17h00)   REQTS 14B/3 MPEG-21           

MPEG-21 FID/DII Issues (16h00 - 16h30) MDS Room 14A/8MDS Plenary  

MPEG-21 DID Comments (09h00 - 10h00) ER/DIA Issues (16h30 - 17h00) MDS Room 14A/8MDS Plenary  

Plenary MDS and Reports of BoG (16h00 - 17h00)   MDS Room 14A/8MDS Plenary  

Review of Output documents, AHGs, CEs, DoC, Std (17h00- 18h00+++)   MDS Room 14A/8             Friday Morning (09h00-13h00)        

Wrapping up (09h00 - 13h00)   MDS Room 14A/8MDS Plenary  

 Approval of resolutions, AHGs and Output documents      

         Friday Afternoon (14h00-21h00)        

MPEG Plenary   Plenary roomMPEG Plenary  

Contact: Ian S Burnett        x        

96

Annex 8Video and JVT

Source: Jens-Rainer Ohm, Gary Sullivan (Video), Miroslaw Bober (MPEG-7 Visual)

1 IDCT in MPEG Standards

The voting on the FCD ballot of 11172-6, which specifies accuracy requirements for IDCT implementation in MPEG video standards, had a favorable outcome (as reported in M12183), with all votes favorable (1 abstention), and with the following comments made:

1. Germany approves document SC 29N6586 but has the following comment: "Any changes should be made that would better harmonize the FDIS 11172-6 text with the related upcoming Corrigenda on 11172-2, 11172-4, 13818-2, 13818-4, 14496-2 and 14496-4."

2. The USA also approved the ballot document, but provided the following editorial comments:a. Ensure that the text is complete and correct (and in concordance with Hong Kong

USNB position remarks) for establishing conformance to MPEG video standards in a manner independent of the former IEEE 1180 standard.

b. The title of the standard should be renamed to clarify its scope as a conformance/tolerance requirements specification rather than a specified implementation. (comments to be further refined in AHG activity)

Several refinements of the text of 11172-6 were proposed in M12387 and were reviewed. It was also suggested to consider issues relating to the rounding adequacy of the 64-bit floating-point computations for the equations found in IEEE 1180. This remains an open issue for study. The highest priority in this category of work seems to be to do a very careful job of replacing IEEE 1180 references with appropriate substitute material. The primary result of the work was the issuance of a study text with a plan to promote the eventually-refined FCD to an FDIS in a manner synchronized with the corrigenda that reference it.

In addition to the input contributions toward preparation of three corrigenda, several additional corrigenda are needed to fix the references to ANSI/IEEE 1180 in MPEG conformance specifications in addition to what is found in MPEG video specifications.

Clarification of the title of 11172-6 is advisable to ensure that its title properly reflects its scope.

Considering the applicability of the accuracy specification across several major MPEG video coding standards, it was agreed that the most appropriate action would be to alter the plan to create 11172-6 and instead create a new video tools suite 23002, of which this specification would become part 1. See remarks below regarding N7516 and the creation of ISO/IEC 23002.

Output documents:No. Title TBP Available

11172-2 MPEG-1 Video7289 Text of ISO/IEC 11172-2 DCOR4 No 05/09/09

11172-4 MPEG-1 Conformance 7291 Text of ISO/IEC 11172-4 DCOR1 No 05/09/09

11172-6 MPEG-1 Implementation of IDCT 7292 Study of ISO/IEC 11172-6 FCD Yes 05/07/29

97

13818-2 MPEG-2 Video7293 Text of ISO/IEC 13818-2:2000 DCOR2 No 05/09/09

13818-4 MPEG-2 Conformance 7295 Text of ISO/IEC 13818-4:2004 DCOR1 No 05/09/09

14496-2 MPEG-4 Visual7296 Text of ISO/IEC 14496-2:2004 DCOR2 No 05/09/09

14496-4 MPEG-4 Conformance 7299 Text of ISO/IEC 14496-4:2004 DCOR2 No 05/09/09

2 MPEG-4 Part 2 Corrigenda Items

After finalization of 14496-4:2004/Cor.1 and 14496-4:2004/Amd.1/Cor.1, several profiles of 14496-2 have lost capability of full conformance testing. Furthermore, in the context of the new standard for specification of IDCT accuracy requirements it was found, that conformance testing for the N-bit Profile was never specified, nor would it have been adequately supported for more than 8 bits by the formulations in 14496-2 pointing to the previous IEEE 1180 specification. Therefore, as announced by resolutions of previous meetings, the video subgroup started an action to remove profiles which have lost conformance. According to the present status, there were two profiles for which no conformance testing bitstream exists in 14496-4 (FGS and N-bit Profiles), such that the situation is most severe. The following modifications are being implemented in 14496-2:2004 DCOR2 (N7296): Removal of profiles from codepoint tables. The previous entries are marked as "forbidden",

such that related bitstreams would no longer have normative support. Moving of the associated tools into an informative annex.This procedure also simplifies a cleanup (full removal) which optionally could be done later. Removal is also made for the OBMC tool that was never used in any profile.

3 JVT Report

The Joint Video Team (JVT) of ITU-T Q.6/16 and ISO/IEC JTC 1/SC 29/WG 11 held its 16th meeting during 24-29 July 2005 in Poznan, Poland. The JVT meeting was held under the chairmanship of Dr. Gary Sullivan (Microsoft/USA) and Dr Jens-Rainer Ohm (RWTH Aachen/Germany), and under the associate chairmanship of Dr. Thomas Wiegand (Fraunhofer HHI/Germany) and Dr. Ajay Luthra (Motorola/USA). The JVT meetings opened at approximately 9:30 on Sunday 24 July 2005 and closed at 12:53 on Friday 29 July 2005. Approximately 154 people attended the JVT meetings (as recorded on a sign-in sheet passed at the meeting) and approximately 114 input documents were discussed. The meetings took place under ISO/IEC auspices in a co-located fashion with a meeting of ISO/IEC JTC 1/SC 29/WG 11 (MPEG). The subject matter of these activities consisted of work on video coding.

3.1 JVT Input documentsJVT-P000 List of documents of Poznan meetingJVT-P001 [G. J. Sullivan, J.-R. Ohm, A. Luthra, T. Wiegand] AHG Report: Proj mgmt and errataJVT-P002 [T. Wiegand, K. Suehring, A. Tourapis, K. P. Lim] AHG Report: JM text and ref softJVT-P003 [T. Suzuki, L. Winger] AHG Report: Bitstreams & conformanceJVT-P004 [J. Ridge, U. Benzler] AHG Report: SVC core experimentsJVT-P005 [G. Cook] AHG Report: JSVM s/W and new func. integ.JVT-P006 [J. Reichel, H. Schwarz, M. Wien] AHG Report: JSVM & WD textJVT-P007 [G. J. Sullivan] AHG Report: Spatial scalability resampling

98

JVT-P008 [Y.-K. Wang] AHG Report: Err resil test cond's & appsJVT-P009 [M. Wien, H. Schwarz] AHG Report: Coding eff & JSVM perf test condJVT-P010 [T. Suzuki] AHG Report: Study of 4:4:4 functionalityJVT-P011 [G. J. Sullivan, J.-R. Ohm, A. Luthra, T. Wiegand] Meeting report of Busan meetingJVT-P012 [S. Sun] CE10Pt1: Direct interpolation for upsamplingJVT-P013 [S. Sun] CE02Pt2: Deblocking filter for I_BL Spatial SVCJVT-P014 [H. Schwarz, D. Marpe, T. Wiegand] Hierarchical B PicturesJVT-P015 [G. J. Sullivan, J. Lou, M.-T. Sun] Adaptive quantization yielding 10 dB gainJVT-P016 [Y.-L. Lee] Lossless intra coding for improved 4:4:4JVT-P017 [H. Yu, L. Liu] Advanced 4:4:4 profileJVT-P018 [E. Francois, J. Vieron, S. Sun] CE10: Report of resultsJVT-P019 [J. Vieron, E. Francois] CE10: Unified motion upsampling in extended spat scalJVT-P020 [J. Vieron] CE10Pr1: Cross-verif of JVT-P012 direct upsampJVT-P021 [X. Yi, J. Zhang, N. Ling, W. Shang] Improved & simplified fast ME for JMJVT-P022 [V. Bottreau] CE09: Cross-verif of FT JVT-P030 on SVC quantizationJVT-P023 [V. Bottreau] CABAC variant and options for bit error resilienceJVT-P024 [T. Suzuki] Editor's proposed draft FRExt conformanceJVT-P025 [T. Suzuki] Prop list of changes of AVC conformanceJVT-P026 [A. M. Tourapis, H.-Y. Cheong, P. Topiwala] Fast ME in the JM reference softwareJVT-P027 [Withdrawn] WithdrawnJVT-P028 [J. Reichel, D. Santa Cruz, F. Ziliani] CE01: Improved residual picture buffer managementJVT-P029 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12304) CE05: CE report on quality layersJVT-P030 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12305) CE09: CE report on improved quantizationJVT-P031 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12306) Syntax for FGS pass fractioningJVT-P032 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12307) On high-level syntax for SVCJVT-P033 [Withdrawn] WithdrawnJVT-P034 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12309) CE02: Cross-check of Aachen prop on deblocking QQJVT-P035 [M.-H. Lee] CE01: Verification of VisioWave proposal JVT-P028 for CE1JVT-P036 [M.-H. Lee] CE06: Verification of MSRA proposal JVT-P047JVT-P037 [M.-H. Lee] CE01: Reduced Memory MCTFJVT-P038 [M.-H. Lee] CE01: Picture Duplication ProcessJVT-P039 [Withdrawn] WithdrawnJVT-P040 [M.-H. Lee] Progressive MCTFJVT-P041 [L. Xiong] Reducing enh.-layer directional intra pred modesJVT-P042 [T.-C. Thang, T.-M. Bae, Y.-J. Jung, Y.-M. Ro, J.-W. Kang, H.-C. Choi, J.-G. Kim, J.-W. Hong] CE08: Spatial scalability of multiple ROIsJVT-P043 [T.-M. Bae, T.-C. Thang, D.-Y. Kim, Y.-M. Ro, J.-W. Kang, J.-G. Kim, J.-W. Hong] FMO implementation in JSVMJVT-P044 [H.-C. Choi] Embedded quant. for improved SNR scalabilityJVT-P045 [I.-H. Shin, H.-W. Park] Adapt. upsamp. for spat. scal.JVT-P046 [J.-Z. Xu] Alternative to FGS quant. method of JVT-O054.JVT-P047 [J.-Z. Xu] CE06: Proposed approach to CE06JVT-P048 [W.-S. Kim+] Further experimental results on 4:4:4 codingJVT-P049 [W.-S. Kim+] Study on advanced 4:4:4 codingJVT-P050 [J. Jia] Polyphase downsamp.-based redundant slices

99

JVT-P051 [J. Jia] Imp. coding eff. by polyphase downsamplingJVT-P052 [X. Wang, Y. Bao, M. Karczewicz, J. Ridge] CE06: Simplified update step operation for MCTFJVT-P053 [Y. Bao, M. Karczewicz, X. Wang, J. Ridge] CE09: Improved quantization for FGS codingJVT-P054 [J. Ridge, M. Karczewicz, Y. Bao, X. Wang] CE11: Core experiment results for FGS reorderingJVT-P055 [Y. Bao, M. Karczewicz, X. Wang, J. Ridge] CE12: Weighted pred. FGS coding closed-loop P framesJVT-P056 [J. Ridge, M. Karczewicz, Y. Bao, X. Wang] Variable length codes for SVCJVT-P057 [X. Wang, Y. Bao, M. Karczewicz, J. Ridge] Implementation of close-loop coding in JSVMJVT-P058 [Withdrawn] WithdrawnJVT-P059 [H. Schwarz, D. Marpe, T. Wiegand] Comparison of MCTF and closed-loop hierarch. B pics.JVT-P060 [X. Wang] CE06: Verif. of Samsung prop.JVT-P061 [Y.-K. Wang, M. M. Hannuksela] Signaling of scalab. info.JVT-P062 [Y.-K. Wang, Y.-L. Bao] Indication of non-required pics.JVT-P063 [P. Yin, J. Boyce, P. Pandit] CE07: Proposal for CE07 enh. layer intra pred.JVT-P064 [P. Yin, J. Boyce, P. Pandit] Weighted prediction for SVCJVT-P065 [P. Yin, J. Boyce, P. Pandit] Weighted prediction for SVC MCTF update stepJVT-P066 [P. Pandit] CE07: Verif. Sharp proposal JVT-O010JVT-P067 [W. Choi, J. Yang, B. Jeon] CE03: CE report coding eff. improvement of CABACJVT-P068 [W. Choi, J. Yang, B. Jeon] CE03: Verif. Samsung prop. JVT-P089JVT-P069 [A. Segall, S. Lei] CE07: Prop. for adaptive upsamp. spatial scalab.JVT-P070 [A. Segall] CE07: Verif. of JVT-P063 enh. layer intra pred.JVT-P071 [A. Segall] Verif. of JVT-P064 weighted pred.JVT-P072 [S. Bandyopadhyay, Z. Wu, P. Pandit, J. Boyce] Frame loss concealmentJVT-P073 [Withdrawn] WithdrawnJVT-P074 [S.-W. Park, J.-H. Park, D.-H. Yoon, B.-M. Jeon] CE10: Verif. of Thomson proposal JVT-O041/O042JVT-P075 [K.-H. Lee] CE04: Report of results of CE04JVT-P076 [K.-H. Lee] Multi-layer weighted predictionJVT-P077 [D.H. Yoon, B.-Y. Jeon, J.-H. Park, S.-W. Park] Complexity reduction by simplified chroma update stepJVT-P078 [D.H. Yoon, B.-Y. Jeon, J.-H. Park, S.-W. Park] Clarification of intra_base_flagJVT-P079 [D.H. Yoon, B.-Y. Jeon, J.-H. Park, S.-W. Park] Key picture indication in AVC compatible base layerJVT-P080 [S. Jeong, M. Park, G. Park, K. Kim] Flag bit in SEI for adaptive GOP structureJVT-P081 [T. Kimoto] CE04: Verification of QQJVT-P082 [T. Kimoto] Multi-loop decoding for unrestricted temporal decomp.JVT-P083 [Y. Yamada, S. Sekiguchi, Y. Isu, K. Asai, T. Murakami] Consideration on intra coding eff. of High 4:4:4 pr. (not Sunday afternoon)JVT-P084 [W.-J. Han] CE06: Core experiment report on update stepJVT-P085 [W.-J. Han] Smoothed ref. pred. for single-loop decodingJVT-P086 [W.-J. Han] Distribution-based inverse quantizationJVT-P087 [W.-J. Han] PFGS with bilinear interpolation filterJVT-P088 [T. Wedi, S. Kondo, H. Saito] 4:4:4 Intra-only codingJVT-P089 [B.-K. Lee] CE03: Report on symbol pred. of resid. pred. flagJVT-P090 [B.-K. Lee] Adaptive FGS processJVT-P091 [B.-K. Lee] CE03: Verification of SKKU proposal JVT-P067

100

JVT-P092 [M. Boltz, U. Benzler] Verif of HHI results JVT-P059 on closed-loop decodeJVT-P093 [M. Beermann, K. Hanke, M. Wien] CE02: Deblock. Filt. Adaptation for open-loop codingJVT-P094 [M. Beermann, M. Wien] CE02: Cross-check of Sharp JVT-P013 on deblockingJVT-P095 [T. Rusert, M. Wien] Locally adaptive quantiz.: concept and first resultsJVT-P096 [W.-J. Han] CE06: Verif of Nokia proposal JVT-P052JVT-P097 [G. H. Park, K. I. Lee] Advanced FMO for ROI in SVCJVT-P098 [J. Reichel, D. Santa Cruz] Verification of Panasonic contribution and encoder improvementJVT-P099 [Withdrawn] Withdrawn JVT-P100 [P. Amon, A. Hutter, T. Rathgen] High-level syntax for SVCJVT-P101 [Withdrawn] WithdrawnJVT-P102 [H. Yu] Artifacts seen in some Viper 4:4:4 sequencesJVT-P103 [A. Rodriguez] Sample aspect ratio supportJVT-P104 [J.-H. Park, D.H. Yoon, S.-W. Park, B.-M. Jeon] CABAC context model for base_mode_flagJVT-P105 [J. Reichel, D. Santa Cruz, F. Ziliani] Is there really a need for a normative update process ?JVT-P106 [Z. Lu] CE8 report: Perceptual Region-of-interest (ROI) based Scalable Video CodingJVT-P107 [Z. Lu] CE8: Cross-check of Spatial scalability of multiple ROIsJVT-P108 [J. Ridge] Verification of CE5 results for France TelecomJVT-P109 [M. Zhou] Verification of JVT-P053JVT-P110 [Withdrawn] WithdrawnJVT-P111 [G. Sullivan] Decoder Inference of Optimal Reconstruction Values for DZ+UTQ QuantizationJVT-P112 [S. Pateux, Y.-K. Wang, P. Amon] Agreed Syntax for FGS pass fractioningJVT-P113 [Y.-K. Wang, T. C. Thang] Agreed Signaling of scalab. info. SEI messageJVT-P114 [J. Reichel, S. Jeong] Simplified syntax for AGS

3.2 Major JVT output documentsBelow are listed the JVT output documents related to AVC as provided to the WG 11 parent body:

No. Title TBP Available14496-4 MPEG-4 Conformance

7306 Disposition of Comments on ISO/IEC 14496-4:2004 FPDAM9 No 05/07/297307 Text of ISO/IEC 14496-4:2004 FDAM9 No 05/08/19

14496-5 MPEG-4 Reference Software 7308 Disposition of Comments on ISO/IEC 14496-5:2001 FPDAM8 No 05/07/297309 Text of ISO/IEC 14496-5:2001 FDAM8 No 05/09/15

14496-10 MPEG-4 Advanced Video Coding

7310 Working Draft 3 of ISO/IEC 14496-10:200x/AMD1 Scalable Video Coding

No 05/07/29

7311 Joint Scalable Video Model (JSVM) 3 No 05/08/297312 JSVM 3 Software No 05/09/307313 Justification of New 4:4:4 Profiles No 05/07/29

3.3 JVT internal output documentsBelow are listed the JVT internal output documents other than SVC core experiment descriptions.JVT-P205 Testing conditions for coding efficiency proposals [05/08/01]JVT-P206 Test conditions for error resiliency proposals [05/07/29]

101

JVT-P207 Requirements and guidelines for normative text submissions to the editors [05/07/29]

3.4 SVC core experiment output documents

Below are listed the JVT SVC core experiment description documents.

JVT-P301 [W.-J. Han, et al.] CE1: update step [based on JVT-P047, JVT-P052, JVT-P084, JVT-P077] (Participants: Samsung, MS, Nokia, LG, HHI, FT, Panasonic, GE, RWTH) [05/07/29]Document has been presented. More discussion over the e-mail reflector is required to refine the CE description.JVT-P302 [K.-H. Lee, et al.] CE2: inter-layer motion prediction [based on JVT-P075] (Participants: Samsung, LG, Nokia, HHI, LM, Sharp, Thomson) [05/07/29]Anchor generation has been discussed. Anchors will be RD optimized wrt to use of co-located MBs and spatial vs. temporal direct mode. Document has been presented. More discussion over the e-mail reflector is required to refine the CE description.

JVT-P303 [T. Kimoto, et. al] CE3: inter-layer texture prediction [based on JVT-P045, JVT-P051, JVT-P082, JVT-P085, JVT-P090] (Participants: NEC, Samsung, KAIST, Nokia, Sejong U., ETRI, HHI, Sharp, Panasonic, GE, LG, Thomson, MS) [05/07/29]Document has been presented. More discussion over the e-mail reflector is required to refine the CE description.

JVT-P304 [Z. Lu, et. al] CE4: ROI [based on JVT-P042, JVT-P106, JVT-P097] (Participants: I2R, Nokia, ETRI, Kyunghee U., ICU, SKKU, Panasonic, HHI) [05/07/29]Document has been presented. More discussion over the e-mail reflector is required to refine the CE description.

JVT-P305 [S. Pateux, et. al] CE5: quantization [based on JVT-P030, JVT-P053, JVT-P086, JVT-P111] (Participants: FT, Nokia, Samsung, HHI, RWTH, Sharp, Panasonic, Thomson, Siemens, MS) [05/07/29]Document has been presented. More discussion over the e-mail reflector is required to refine the CE description.

JVT-P306 [E. Francois, et. al] CE6: extended spatial scalability [based on JVT-P019] (Participants: Thomson, Nokia, Sharp, FT, HHI, GE, LG) [05/07/29]Document has been presented. More discussion over the e-mail reflector is required to refine the CE description.

JVT-P307 [Y. Bao, et. al] CE7: low-delay FGS [based on JVT-O054, JVT-P046, JVT-P087] (Participants: Nokia, GE, Panasonic, LM, HHI, MS, Samsung) [05/07/29]Document has been presented. More discussion over the e-mail reflector is required to refine the CE description.

JVT-P308 [J. Ridge, et. al] CE8: CAVLC [based on JVT-P056] (Participants: Nokia, HHI, GE, TI, Panasonic, SKKU) [05/07/29]Document has been presented. More discussion over the e-mail reflector is required to refine the CE description.

102

3.5 JVT Administrative topics3.5.1 Administrative documentsJVT-P000 List of documents of Poznan meetingJVT-P001 [G. J. Sullivan, J.-R. Ohm, A. Luthra, T. Wiegand] AHG Report: Proj mgmt and errataDocument contains an introduction to the JVT and its work. This is recommended reading for those not familiar with the JVT. Also found in the document are recent reports of AVC specification errata.JVT-P002 [T. Wiegand, K. Suehring, A. Tourapis, K. P. Lim] AHG Report: JM text and ref softDocument reviews state of work on reference software and JM algorithm description text.JVT-P003 [T. Suzuki, L. Winger] AHG Report: Bitstreams & conformanceDocument reviews state of work on bitstream exchange interoperability work and conformance specification.JVT-P004 [J. Ridge, U. Benzler] AHG Report: SVC core experimentsDocument reviews state of "core experiment" collaborative work on algorithm testing for SVC project.JVT-P005 [G. Cook] AHG Report: JSVM s/W and new func. integ.AHG report has been noted. JSVM 2.0 software is available and has been approved by the JVT.JVT-P006 [J. Reichel, H. Schwarz, M. Wien] AHG Report: JSVM & WD textReport has been noted. JSVM 2 and WD 2 text have been approved. New procedure: Use Annex S in JSVM for all normative elements. This can be copied into the WD right away.

JVT-P007 [G. J. Sullivan] AHG Report: Spatial scalability resamplingReviews state of work on resampling. Email conversation held on reflector during the meeting containing relevant information has been added in an "r1" version of the document.JVT-P008 [Y.-K. Wang] AHG Report: Err resil test cond's & appsReport has been noted. Testing conditions have been adopted by the JVT as described in JVT-P206 subject to further refinement until 45 days prior to the JVT Nice meeting.JVT-P009 [M. Wien, H. Schwarz] AHG Report: Coding eff & JSVM perf test condDetailed Results provided for JSVM2 single layer, spatial scal., SNR scal. and combined scalability under new testing rate points as proposed earlier on the reflector. The JVT agrees to adopt these testing conditions for future benchmarking of SVC development and core experiments as described in JVT-P205.JVT-P010 [T. Suzuki] AHG Report: Study of 4:4:4 functionality

Test conditions recommended. JM 9.6 used. Display available here in Poznan may not be suitable for subjective evaluations. QP 12, 18, 24, 30 used – also optional 6, 36, 42. Should try to harmonize as appropriate with other test conditions work (e.g., not using QP values spaced apart by increments of 6).Relevant inputs 16, 17, 48, 49, 83, 88, 102.JVT-P011 [G. J. Sullivan, J.-R. Ohm, A. Luthra, T. Wiegand] Meeting report of Busan meeting

3.5.2 IPR policy reminderParticipants were reminded of the IPR policies established by the parent organizations of the JVT and were referred to the parent body web sites for further information.

Participants were particularly reminded of the need to supply a completed JVT IPR status reporting form in all technical proposals for normative standardization.

103

3.5.3 Late documentsNo objections / objections were voiced to the consideration of the late documents. Documents listed in this report in italics and with a "-L" suffix to their document numbers were classified as late.JVT-P017-L [H. Yu, L. Liu] Advanced 4:4:4 profileJVT-P022-L [V. Bottreau] CE09: Cross-verif of FT JVT-P030 on SVC quantizationJVT-P024-L [T. Suzuki] Editor's proposed draft FRExt conformanceJVT-P033-L [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12308) CE05: Cross-check of Nokia prop on quality layers QQJVT-P046-L [J.-Z. Xu] Alternative to FGS quant. method of JVT-O054.JVT-P048-L [W.-S. Kim+] Further experimental results on 4:4:4 codingJVT-P049-L [W.-S. Kim+] Study on advanced 4:4:4 codingJVT-P053-L [Y. Bao, M. Karczewicz, X. Wang, J. Ridge] CE09: Improved quantization for FGS codingJVT-P054-L [J. Ridge, M. Karczewicz, Y. Bao, X. Wang] CE11: Core experiment results for FGS reorderingJVT-P055-L [Y. Bao, M. Karczewicz, X. Wang, J. Ridge] CE12: Weighted pred. FGS coding closed-loop P framesJVT-P056-L [J. Ridge, M. Karczewicz, Y. Bao, X. Wang] Variable length codes for SVCJVT-P057-L [X. Wang, Y. Bao, M. Karczewicz, J. Ridge] Implementation of close-loop coding in JSVMJVT-P063-L [P. Yin, J. Boyce, P. Pandit] CE07: Proposal for CE07 enh. layer intra pred.JVT-P067-L [W. Choi, J. Yang, B. Jeon] CE03: CE report coding eff. improvement of CABACJVT-P070-L [A. Segall] CE07: Verif. of JVT-P063 enh. layer intra pred.JVT-P102-L [H. Yu] Artifacts seen in some Viper 4:4:4 sequencesJVT-P105-L [J. Reichel, D. Santa Cruz, F. Ziliani] Is there really a need for a normative update process ?JVT-P106-L [Z. Lu] CE8 report: Perceptual Region-of-interest (ROI) based Scalable Video CodingJVT-P107-L [Z. Lu] CE8: Cross-check of Spatial scalability of multiple ROIsJVT-P108-L [J. Ridge] Verification of CE5 results for France Telecom

3.6 JVT Scalable video coding work

3.6.1 CE1 & related docs: Memory management for MCTF processJVT-P028 [J. Reichel, D. Santa Cruz, F. Ziliani] CE01: Improved residual picture buffer managementThe special case where a single residual buffer is used is investigated. The decoding update process is modified as follows: Each update process using N reference pictures is split into N update processes using only a single input picture each. This algorithm is called the split-update. Using the algorithm reduces the decoding latency, and simplifies the decoding process. This contribution shows that this algorithm can be introduced into the JSVM without any degradation of the compression performances.

MCTF decoding requires storage of residual pictures until the inverse prediction step, number cannot be controlled by num_ref_frames, higher precision required. Solutions: Constraints on MCTF structure (proposed in JVT-O009) or decode with fixed number of residual picture buffer as proposed here. Modification of the encoder could be done, but is not necessary (in some cases slight

104

improvements, in some cases slight degradations as compared to JSVM, with 1%, when encoder is unchanged). Difference most probably caused by different effect of adaptive update.

JVT-P040 [M.-H. Lee] Progressive MCTFAn approach called progressive motion-compensated temporal filtering is described. It modifies the MCTF structure such that MCTF can be performed in a progressive and straightforward manner. It affects the order of processing the prediction and update steps and the choice of reference pictures for each step. The prediction and update steps are interleaved such that the low-pass pictures are partially updated immediately with each high-pass picture in a progressive manner and the partially updated low-pass pictures are used for subsequent prediction steps. The progressive MCTF is claimed to simplify the inverse MCTF process, eliminate irregular complexity peaks, lower decoding delay and lower memory requirements, compared to the conventional MCTF in the JSVM, without any penalty on coding efficiency.

Similar to JVT-P028 (in fact, the same. Interleave prediction&update progressively instead of waiting for all residual pictures that are involved in a specific update.

This was agreed to be adopted.

JVT-P035 [M.-H. Lee] CE01: Verification of VisioWave proposal JVT-P028 for CE1Verification of JVT-P028. The binaries are compiled and built from the modified software provided through JVT-P028 and the results are generated using the same set of configurations provided through JVT-P028. The results obtained are similar to JVT-P028 results.

Cross-check 100% when run on same platform. Most probably a memory initialisation problem still existing in JSVM2 software.

JVT-P037 [M.-H. Lee] CE01: Reduced Memory MCTFAn approach called 8Bit_Separate is described to reduce memory requirement of storing a high-pass residual picture. It uses the MSB 8-bits of decoded residual of a high-pass picture for modified 8-bits inverse update operation. An alternative solution called 8Bit_Saturate is evaluated too. It uses the saturated 8-bits of decoded residual of a high-pass picture for inverse update operation by saturating decoded residual to 8-bits (-127 to 128). Complexity reduction, minor loss occuring in some cases.

It was agreed to adopt clipping of the reference signal prior to update to be represented by 8 bits per sample.

JVT-P038 [M.-H. Lee] CE01: Picture Duplication ProcessFor the current Joint Scalable Video Model, picture duplication is required by the decoder for the lowpass picture of the current GOP. To ensure conformance between encoded bitstream and a decoder having a finite amount of memory, the memory used by the duplicated picture has to be controlled, from the instance it is created, used and finally removed. The current methods used by the SVC WD and the JSVM software for handling the duplicated picture are examined. Some issues are caused by the original buffer and the duplicated picture buffer having the same index. It is proposed that the duplicated picture is assigned a long-term index, to differentiate it from the original picture. This is done using an MMCO command proposed previously, with an addition of marking the duplicated picture as “used for long-term reference” and assigned a long term index.

105

Problem in buffer management with SNR scalability (different decoding processes for key pictures) is recognized, but doubts are raised whether the proposed solution is general enough. Get together with the editors to discuss solution.

Problem is only solved for CGS and 2 layers only. Agreed.

It may also need to be solved for more than 2 CGS layers but that is not clear. No solution when FGS is involved.Further work is encouraged to address the open problems.

Consider in CE.

JVT-P098 [J. Reichel, D. Santa Cruz] CE01: Verif of Panasonic proposalsVerification of JVT-P037 contribution to CE1 on “Memory Management for MCTF process”. Proposes minor encoder modifications to avoid artifacts in the decoded pictures. JVT-P037 proposes two solutions to reduce the precision of the buffers used to store inter MBs to 8 bits: split the storage into two buffers or simply saturate (i.e. clip) the residuals to 8 bits. The fist proposal has been fully verified using source code and description provided by Panasonic. The second proposal is claimed to have the potential to create visible artifacts in the decoded pictures. In addition to verifying the results in JVT-P037, an encoder modification is proposed that completely avoids such artifacts with a small degradation of the rate-distortion performance. The proposed modification consists in constraining the mode decision process so as to forbid inter modes for the current macroblock that would lead to saturation of the residual signal at the decoder.

Verifies JVT-P037 and modifies the encoder by constrained mode decision such that artifacts are avoided. Loss in RD performance is up to 0.16 db (Mobile), 0.12 (Bus), negligible in other sequences. Encoder is becoming more complex.

JVT-P057-L [X. Wang, Y. Bao, M. Karczewicz, J. Ridge] Implementation of closed-loop coding in JSVMInformation on closed-loop coding implementation in JSVM1.0. While the implementation of closed-loop for discrete layer coding is trivial in JSVM, it is reported that is takes some extra efforts to optimize the performance of the codec for combined scalabilities, especially the FGS scalability. For the results presented in this contribution, how the FGS layer of an anchor frame is coded is not changed. The focus of this contribution will be on close-loop FGS layer coding of high-pass frames.

Comparison: Open loop (original for ME and MC at encoder), "partial" closed loop with all-FGS for MC at encoder, "true" closed loop with partial-FGS or no-FGS for MC at encoder. Also non-originals for ME in latter cases. Impact on quantizer and entropy coder should be observed. Gain up to 0.5 dB for the case of "true" closed loop vs. open loop. No change of decoder, multiple loop at encoder.

JVT-P059 [H. Schwarz, D. Marpe, T. Wiegand] Comparison of MCTF and closed-loop hierarch. B pics.The coding efficiency of MCTF-based coding and a simple closed-loop coding with hierarchical B pictures for single-layer coding as well as spatial, SNR, and combined scalability. For all test scenarios, the coding efficiency with hierarchical B pictures and a closed-loop control is claimed to be higher or similar to that of MCTF-based coding for nearly all test sequences and resolutions. Only for the “City” sequence MCTF-based coding yields a coding efficiency gain of up to 0.5 dB in

106

comparison the closed-loop coding with hierarchical B pictures. A further analysis of these gains for “City” lead to the conclusion, that the gains for the “City” are a result of the temporal noise reduction property of the MCTF. It is claimed to be shown that this noise reduction can also be realized as a pre-processing step and efficiently combined with closed-loop hierarchical B picture coding. Based on the simulation results and the fact that the update steps at the decoder side dramatically increase the complexity, it is propose to remove the update steps from the SVC design.

Results indicate that we can have a comparison benchmark that in most cases is better than MCTF open-loop design. To be discussed in the context of discussions on CE6.

JVT-P092 [M. Boltz, U. Benzler] Verif of HHI results JVT-P059 on closed-loop decode

In JVT-P059, a comparison between the JSVM MCTF coding and closed-loop coding is described. The encoded bit streams from HHI have been received and successfully decoded them using the version of the JSVM described in JVT-P059 (CVS Tag JSVM_1_2_5). The results match exactly those published in JVT-P059.

JVT-P105-L [J. Reichel, D. Santa Cruz, F. Ziliani] Is there really a need for a normative update process ?It is claimed that it is possible to remove the update process from the decoding process without reducing the compression performance of the JSVM. When the update process is performed on the encoder side but not on the decoder side, the decoded video is claimed to be of nearly of the same quality as JSVM decoding with update step for most test sequences. It should be noted that also removing the update process from the encoder will, in this case, reduce the coding performances. The experiments conducted in this contribution seem to show the update process is in fact necessary only on the encoder side, and could be made non-normative.Omitting update step at decoder gives little loss (< 0.1 db) for CIF, but up to 0.4 dB for 4CIF (City, Harbour). Omitting update at both encoder and decoder gives more critical loss. Mainly PSNR in lowpass pictures is decreased. Difference becomes larger when the bitrate increases. Needs more investigation probably also for higher rates; there seems to be some tendency that encoder_update+decoder_noupdate comes below encoder_noupdate+decoder_noupdate.

For efficient coding results, current information indicates that closed-loop coding must be used.Open question with regards to use of the update step within that scenario:

1. Normative decoding feature2. Pre/post filter (non-normative)

The JVT decided to create a core experiment on the update step.Use closed-loop coding without the update step as a reference merged with quality layers.

3.6.2 CE 2 & related docs: Improved deblocking filter settingsJVT-P013 [S. Sun] CE02Pt2: Deblocking filter for I_BL Spatial SVCA simplification of the deblocking filter for macroblocks in I_BL mode in spatial scalable video coding is proposed. By treating the I_BL blocks as inter blocks during boundary strength decision, the deblocking filter operations for I_BL blocks are claimed to be significantly reduced (34% on average) while acceptable visual quality is claimed.Provide more precise results relating to count of specific operations that are saved. Communicate offline with people who still have concerns. If clarified, adopt to JSVM.

The JVT agreed to adopt as described in JVT-P013r1.

JVT-P094 [M. Beermann, M. Wien] CE02: Cross-check of Sharp JVT-P013 on deblocking

107

Verification of JVT-P013. The source of the proposal as well as the bitstreams has been received. The verification has been performed with the compiled source. The RD results as well as the filter statistics reported in JVT-P013 are verified.

JVT-P093 [M. Beermann, K. Hanke, M. Wien] CE02: Deblock. Filt. Adaptation for open-loop codingThe H.264|AVC deblocking filter has been designed and tuned for closed-loop application. The SVC extension is claimed to contain an open-loop structure. This fundamental difference has not yet been considered in the necessary adaptations of the deblocking filter for its use in the scalable extension. The differences that evolve from the open-loop situation is discussed. Proposed modifications to the deblocking filter are compared to a very simple adaptation of the deblocking filter through available slice-header syntax. This comparison includes rate-distortion results as well as visual coding results.

To be considered within open-loop/closed-loop investigation of AHG on Coding eff & JSVM perf test cond.

JVT-P034 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12309) CE02: Cross-check of Aachen prop on deblocking QQ

Verification of JVT-P093.

3.6.3 CE 3 & related docs: Coding efficiency improvement of entropy codingJVT-P067-L [W. Choi, J. Yang, B. Jeon] CE03: CE report coding eff. improvement of CABACThis contribution proposes the adaptive entropy coding scheme by using inter-layer relationship for coding efficiency improvement in response to CE 3. A new context modeling process of CABAC is proposed based on a syntax element of the base layer macroblock. In simulation results, the overall coding gain of the proposed scheme is claimed to be up to 1% in terms of bit saving. The coding gain of the proposed scheme is reported to be increased as the slice size is reduced. In case of mobile application which has a limitation on packet size, the proposed scheme would have a benefit on coding efficiency.

Number of new models: header: 33, residual data: 844Size of new binarization table: 23 symbolsDid not follow new test conditions.

Contribution has been noted. JVT-P091 [B.-K. Lee] CE03: Verification of SKKU proposal JVT-P067Verification of JVT-O067. Due to the tight deadline, the verification results for spatial test condition and Munich test configuration have not been finished. All other sequences are verified and there are no problems. The details are described in the attached excel file. (JVT-P091.xls) For the unfinished sequences, the results until the Poznan meeting will be provided.

JVT-P089 [B.-K. Lee] CE03: Report on symbol pred. of resid. pred. flagAccording to the current JSVM2 implementation, residual_pred_flag is set to one if the base-layer residual is used for predicting the current inter residual. However, if the base-layer residual in the macroblock has no non-zero pixel, the coding of residual_pred_flag is not required. Thus we encode reversed residual prediction flag instead of the residual_pred_flag itself to increase number of '0'. It provides some performance gain when the VLC is used instead of CABAC. This process can be

108

summarized as: 1) Skip coding of residual_pred_flag if base-layer residual has no non-zero pixels, 2) Code reversed residual prediction flag = 1-resiudal_pred_flag

Bit-rate savings around 1-1.4%.

The JVT agreed to adopt, by adding a context model that is conditioned on the CBP value of the base layer.

JVT-P068 [W. Choi, J. Yang, B. Jeon] CE03: Verif. Samsung prop. JVT-P089Verification of JVT-P089 for CE 3. Source code based on the reference software JSVM 2.0 has been provided. It has been checked that the provided source code has no problem and the coded bitstreams are also verified. The proposed scheme is tested under modified Munich test condition and coding performance is checked in terms of bitrate. To obtain the bit reduction ratio of the proposed method, tests were conducted without rate control. The layers are decided by fixed QP value not bitrate.

JVT-P104 [J.-H. Park, D.H. Yoon, S.-W. Park, B.-M. Jeon] CABAC context model for base_mode_flagCABAC can be divided into three modules such as, binarizer, context modeler and binary arithmetic coder. In the current JSVM 2.0, one of the most important modules, the context modeler uses the same syntax elements of neighboring maroblocks to calculate context index increment. For the single layer predictive coding scheme, using the same syntax of the neighboring macroblocks is reasonable. However, regarding the multi-layer scheme, syntax elements of upper layer can obtain the information from the lower layer. If the macroblock of spatial enhancement layer has a base macroblock and the two macroblocks have the same temporal location, their relationship is claimed to be closer than the relationship of neighboring macroblocks. Measurements on the current context model for base_mode_flag based on Palma conditions show the actual percentages. Expanded context models for base_mode_flag incorporating the effect of the macroblock modes of base layer are proposed.

Number of new models: 3Did not follow new test conditions.

Further work is encouraged.

3.6.4 CE 4 & related docs: Inter-layer motion predictionJVT-P075 [K.-H. Lee] CE04: Report of results of CE04Extension of JVT-O058. In JSVM 2.0 the motion prediction method of the picture which has a time synchronized base layer picture at lower temporal resolution (called synchronized picture) cannot be used because the analysis step of the MCTF process estimates the motion field starting from the highest temporal level proceeding through to lower temporal levels. The method proposed in JVT-O058 estimated a motion predictor from the base layer motion for the unsynchronized picture. If the MCTF analysis process can be separated into 2 stages, motion estimation and decomposition stages, then the motion predictor of the unsynchronized picture can be derived from the lower temporal level picture. This motion predictor estimation is similar to temporal direct mode.

Check of temporal direct mode. Test with JM decoder with only 1 sequence.Combine with closed-loop coding.

Continue CE. Specify more optimal anchors.

109

JVT-P081 [T. Kimoto] CE04: Verification of QQVerification of JVT-P075. Only coding a performance check was carried out by running software.

JVT-P082 [T. Kimoto] Multi-loop decoding for unrestricted temporal decomp.Inter-layer MC with multi-loop decoding (MLD) to improve coding performance for SVC is proposed.

Decouples MC structure between base and enhancement layer.Proposes multiple loop decoding and using enhancement layer MVs to create prediction signals using MC in the Base layer.

Create CE on inter-layer texture prediction.

Take into account also impacts on DPB, level definition wrt memory use, Annex C, MMCO.

JVT-P085 [W.-J. Han] Smoothed ref. pred. for single-loop decodingSingle-loop decoding is a very useful technique to reduce the decoding complexity in the multi-layer structure. It is claimed that the restrictions of the intra-base prediction sometimes degrades the performance especially for fast-motion sequences. A new prediction mode based on the residual prediction process is proposed to simulate the intra-base prediction by adding the smoothing function to the prediction signal with consideration of the residual prediction process. A performance improvement for the sequences Football, Crew, and Soccer is claimed.

Add to CE on inter-layer texture prediction..

3.6.5 CE 5 & related docs: Quality layersJVT-P029 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12304) CE05: CE report on quality layersContribution on quality layers as described in JVT-O305. Rate-distortion optimized quality layers are associated to the SVC NAL headers. A gain of up to 0.4 dB is claimed using quality layers with respect to JSVM2.

The non-normative part was adopted by the JVT.

JVT-P108-LL [J. Ridge] Verification of CE5 results for France TelecomNokia reports verification of France Telecom’s results provided in JVT-P029, by decoding the provided bitstreams using the provided binary executables. They report that the highest-resolution PSNR values are in complete agreement. A discrepancy was observed for lower resolutions, but was reportedly almost certainly due to the use of differently downsampled references used in computing PSNR. It was reported that time did not permit Nokia to investigate this further, but that they would be happy to do so if the group felt it to be necessary

3.6.6 CE 6 & related docs: Improvement of update stepJVT-P047 [J.-Z. Xu] CE06: Proposed approach to CE06An alternative update operator is presented. The update method proposed is based on the one proposed in JVT-O030, but further modifications are made to reduce the computational complexity and to eliminate the need of an extra frame buffer used to store the update components in the originally proposed method. The proposed method is implemented in the current SVC reference

110

software JSVM2. Experiments are conducted according to the test conditions set for CE6. It is claimed that on most of the test sequences, the proposed method can achieve better or similar coding performance. The proposed update operator skips the step of inversing the motion field and that of computing update weights, which must be done in the current update operator in JSVM2. Moreover, the proposed method is claimed to significantly reduce the memory requirement.

Concerns raised regarding division by 7 and its integer approximation.

Threshold values adjusted manually. Concerns raised about this.

Complexity relative to JVT-P052 discussed.

JVT-P036 [M.-H. Lee] CE06: Verification of MSRA proposal JVT-P047Verification of JVT-P047. The binaries are compiled and built from the modified software provided by MSRA and the results are generated using the same set of configurations provided by MSRA. The results obtained are similar to MSRA’s results. See attached excel file JVT-P036.xls.

JVT-P052 [X. Wang, Y. Bao, M. Karczewicz, J. Ridge] CE06: Simplified update step operation for MCTFFollow-up of JVT-O015. New ideas are: 1) Update motion vector derivation process is omitted. The reverse direction of motion vectors used in prediction step is used directly as update motion vectors. 2) Outlier motion vectors are checked and excluded from the update step. 3) Energy based adaptive weighting on update signal is replaced with a simple block energy based capping operation to limit the maximum amplitude of update signal.With all these new simplifications, there is no obvious degradation claimed on coding performance. For some test sequences, the coding performance is claimed to be slightly improved.

For 1) "Direct update" replaces derivation of update motion vector for each 4x4 block. For 2) Up to 50% of motion vectors excluded from update process, so in general less update is made. Minor impact on PSNR. Comparison made against the simplified version (with simple interpolation method). Interpolation is made on basis of 16x16 blocks in cases of 16x16 block mode.

Breakout to identify commonalities/differences between JVT-P047 and JVT-P052.

The JVT agreed to adopt a unified solution (JVT-P052 and JVT-P047) as described in JVT-P052r1.

JVT-P096 [W.-J. Han] CE06: Verif of Nokia proposal JVT-P052

Verification of JVT-P052. By using the provided binaries and configurations files, results in JVT-P052 can be reproduced. However, due to the tight deadline, the verification results for Crew, Harbour, and Crew sequences are nit finished. All other sequences are verified. The details are described in the attached excel file. For the unfinished sequences, results are promised for the Poznan meeting.

Verification has been finished, no problem found.

JVT-P084 [W.-J. Han] CE06: Core experiment report on update stepIt is well-known that the open-loop MCTF with update-step structure provides a state-of-the art performance as well as the excellent scalability feature although most standardized video codecs

111

utilize the closed-loop concept. In this contribution, the various properties between closed-loop and open-loop framework including the single-layer and multi-layer tests are investigated. Furthermore, a new scheme for utilizing both good properties of the closed-loop scheme and the update-step by combining the closed-loop prediction and the switched update-step technique is proposed. From the test results of combined scalability configuration, it can be concluded that the new scheme can combine the closed-loop prediction and the update-step successfully and improve the PSNR values up to 0.8 dB.

Up to 15% savings in bit rate by update step in open-loop. Tool1: Closed-loop prediction also gives gain if update step is used. In base layer, closed loop HB with closed-loop ME shows best performance. In enhancement layer, usage of update step gives additional gain (up to 0.44 dB). Tool2: Performing prediction step first in the decoder allows to avoid mismatch. Gives up to 3 dB gain at small QP (8) in the PSNR range of 40 dB and above. For higher QP value, update step can be skipped at the decoder (QP > 24: no PSNR drop). Would allow to apply complexity scalability at the decoder. Combination of the two tools gives typically 0.2-0.3 dB over JSVM2. Update step gives less PSNR fluctuations. For inter-layer motion prediction, closed loop ME gives sometimes worse results (less coherence between base and enhancement).

Further investigation in CE recommended to find out benefits of update at higher fidelity ranges.

JVT-P060 [X. Wang] CE06: Verif. of Samsung prop.Verification of JVT-P084. Both the encoder and decoder (binary executable) were provided. Using the configuration files provided in JSVM2, sequences were re-encoded, decoded and PSNR values were computed against original at each target bitrate on all the test sequences. The PSNR values from the verification test match those provided in JVT-P084.

JVT-P077 [D.H. Yoon, B.-Y. Jeon, J.-H. Park, S.-W. Park] Complexity reduction by simplified chroma update stepThe update step for the chroma signal is simplified by quantizing chroma update motion vector to integer-pel accuracy. It is claimed that this simplification does not introduce any noticeable degradation but the overall complexity of update step is reduced because fractional pixel interpolation can be avoided. Besides that, skipping of chroma update is also proposed, and this feature could be controlled by using a syntax element disable_chroma_update_flag.

Method 1: Simplification by using integer MV, only small loss in PSNR (<0.1 dB); method 2: Skipping of chroma update entirely: loss of 0.5-1 dB. Comparable to other results reported on the entire omission of the update step, but questionable why it should be best strategy to omit it only for luma. Further discuss in definition of CE with the other participants.

For further consideration of CE6 results, subjective comparison will be performed for- JSVM2- Hierarchical B frames, closed loop- P047- P052- P084

Tests with 4 people did not show any difference between the various methods.

Continue CE.

3.6.7 CE 7 & related docs: Enhancement layer intra predictionJVT-P063-L [P. Yin, J. Boyce, P. Pandit] CE07: Proposal for CE07 enh. layer intra pred.

112

The technical description and experimental result in response to CE7 are provided, which as an update of JVT-O053. Unlike JVT-O053, the new mode I_BLS, which enables spatial intra prediction for the enhancement layer residue, is added as an additional macroblock prediction mode, so that I_BL mode and I_BLS mode can be selected on macroblock basis. Compared to JVT-O053, the new proposal is claimed that it can still maintain the gain at the higher bitrate, while also achieve some gain at lower bitrate.

Bitrate savings of 1.28% for QP 24-36Bitrate savings of 3.53% for QP 8-20

Contribution has been noted.

JVT-P070-L [A. Segall] CE07: Verif. of JVT-P063 enh. layer intra pred.Verification of JVT-P063. Bit-streams and a binary decoder have been provided. It was verified that the bit-streams are decodable and bit-rates and PSNR values were measured. The rate and PSNR information is enclosed in a supporting document.

JVT-P069 [A. Segall, S. Lei] CE07: Prop. for adaptive upsamp. spatial scalab.An approach to adaptive upsampling. The description is based on JVT-O010, which proposes a technique to select filters on a block-by-block basis and utilizes a four-channel filterbank. Filter selection is performed at the encoder and signaled to the decoder. For the case of IntraBL coded blocks, a coding improvement of up to 0.6dB is claimed when compared to the current JSVM and encoded with a PSNR of 38dB. Coding improvements of 1.7dB are claimed for high-rate conditions.

PSNR gains at 40dB: 0-0.2dB for most sequences and 0.46dB for Mobile.Gains are consistent over the two (probably different) versions of the downsampling mechanism.

Question: How could down-sampling be changed to obtain some of the gains?Not clear how some of the sequences have been created. Are there phase shifts.Size of downsampling filters? 13 taps 4CIF->CIF, CIF->QCIF, 9 taps?Markus Beermann asked to provide information on how the originals in our experiments have been generated.

Question on level of adaptation: MB, picture, sequence. Said to be tried and not working.

Contribution has been noted.

The JVT decided to create an AHG on re-sampling [Chair: Gary Sullivan].

JVT-P066 [P. Pandit] CE07: Verif. Sharp proposal JVT-O010Verification of JVT-P069. Related to the direct texture upsampling method.

JVT-P078 [D.H. Yoon, B.-Y. Jeon, J.-H. Park, S.-W. Park] Clarification of intra_base_flagProposal is about I_BL (intra base) macroblock, which uses the scaled and reconstructed base layer signal as the predictor. Two flags are used to indicate I_BL macroblock type; base_mode_flag and intra_base_flag. The usage of intra_base_flag in the current JSVM is claimed to be inappropriate for the single-loop decoding case. Contains a proposed clarification of the intra_base_flag syntax with regards to inter-layer prediction constraint.

113

In single-loop decoding, when Intra_BL prediction is not possible, the flag that indicates the use of Intra_BL prediction should not be present. The JVT agreed to adopt.

JVT-P041 [L. Xiong] Reducing enh.-layer directional intra pred modesIn addition to H.264/MPEG4-AVC Intra prediction maccroblock types in the JSVM2 enhancement layer Intra prediction and the I_BL model is added. For I_BL, the corresponding encoded macroblock information in layer k can be used to process the Intra prediction of current macroblock in the layer k+1. By using I_BL, on some occasions, it is claimed that better efficiency of prediction and encoding can be achieved.

Create an AHG on enhancement slice complexity reduction [L. Xiong]

3.6.8 CE 8 & related docs: Region-of-interest codingJVT-P042 [T.-C. Thang, T.-M. Bae, Y.-J. Jung, Y.-M. Ro, J.-W. Kang, H.-C. Choi, J.-G. Kim, J.-W. Hong] CE08: Spatial scalability of multiple ROIsThe functionality of ROI spatial scalability is addressed. ROI can be encoded by different quality and/or resolution. ROI can be overlapping, duplicate encoding to be avoided. Border extension necessary, e.g. for upsampling filtering of an ROI. Exp. 1: Bitrate saving according to size; Exp.2.: Bitrate saving according to OR size; Exp. 3: Bitrate saving according to "virtual part". Syntax element proposed for signaling number of ROIs, and decoding process for boundary handling by the virtual part.

Hint is made that this might better be done non-normatively. Boundary filtering could be done after decoding, encoder would need to take care that no MV points outside the ROI. FMO is already capable to handle these issues. "Virtual part" could be implemented by skipping macroblocks. Further study encouraged.

JVT-P043 [T.-M. Bae, T.-C. Thang, D.-Y. Kim, Y.-M. Ro, J.-W. Kang, J.-G. Kim, J.-W. Hong] FMO implementation in JSVMFMO(Flexible Macroblock Ordering) and ASO(Arbitrary Slice Order) are already existing tools. Currently, these tools are not supported by both encoder and decoder of reference software of SVC, JSVM2.0. To support these functionalities, FMO and ASO have been implemented in the JSVM 1.0 and the encoding procedure is described.

Get in contact with software integrator. Consider in new integration plan for JSVM software.

JVT-P097 [G. H. Park, K. I. Lee] Advanced FMO for ROI in SVCAn Advanced Flexible Macroblock Ordering (A-FMO) method is proposed with the aim of improved picture quality of the ROI region in which the cyclic-FGS coding method is used. The concept of the virtual rings on the FMO is proposed and the number of recursive scan frequency (RSF) for each virtual ring is controlled. By changing the RSF on each virtual ring, image information of the blocks of each virtual ring can be differently encoded and decoded according to the priority of the user interests, therefore the picture quality of the blocks assigned into the ROI region can be enhanced in comparison with the picture quality of the blocks located in the left-over region. The proposed method uses the box-out scan with introducing the virtual ring concept and applies the different number of scanning for coding to the blocks belong to virtual rings to accomplish the enhancement of the ROI without using the conventional bit-shift process.

Ring/spiral sequence only for FGS refinement. Further work needed showing the possible advantages of the technique as compared to "conventional" FMO.

114

JVT-P106-L [Z. Lu] CE8 report: Perceptual Region-of-interest (ROI) based Scalable Video CodingReports progress in CE8 test 1. Changing ROI are implemented and introduced in the report. Subjective and objective comparison between JSVM1 and perceptual ROI-based JSVM1 are to be presented in the report. Because the subjective viewing test is arranged to hold on Saturday (23rd July 2005) afternoon, a revised version with the comparison on subjective quality will be uploaded later.

Implementation by using existing tools (quantizer update, motion refinement). Some overhead, but not quantified. Presently no concrete proposal how this overhead may be avoided.

JVT-P107-L [Z. Lu] CE8: Cross-check of Spatial scalability of multiple ROIsVerification of CE8 test 1, JVT-P042. The source of the proposal as well as the executable code and configuration files has been received. The bitrate reported in JVT-P042 are verified.

ROI to be further investigated in CE. Use FMO as present way of implementation as an anchor for comparison. To define common testing conditions, the center part of the sequence should be enhanced by ROI, and the amount of enhancement (e.g. delta-QP) must be commonly used. Any future proposals on ROI shall use these conditions.

3.6.9 CE 9 & related docs: Improved quantizationJVT-P030 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12305) CE09: CE report on improved quantizationAn adaptive reconstruction offset has been tested as well as embedded quantization intervals as proposed in JVT-O060. Using new test conditions, some gain is claimed in SNR scenarios (up to 0.23dB on luminance and 0.31dB on chromas for higher rates). For some sequences such as Mobile no specific gain is observed. When considering combined scalability scenarios, gain is only observed at the lowest resolution. This gain is not passed on to higher resolution.

Continue CE.

JVT-P022-L [V. Bottreau] CE09: Cross-verif of FT JVT-P030 on SVC quantizationVerification of JVT-P030. The test conditions from JVT-P030 were used.

JVT-P053-L [Y. Bao, M. Karczewicz, X. Wang, J. Ridge] CE09: Improved quantization for FGS codingNokie reportedly verified the simulation results presented in JVT-P053 entitled “CE09: Improved quantization for FGS coding results”. All bitstreams were decoded (in total 256 bitstreams for eight test sequences, 16 truncation points, and two codecs), PSNRs were computed against the original sequences, and the bit-rates were measured according to the file size of bitstreams. The results were reportedly identical to those presented in spreadsheet JVT-P053.xls. The executable decoders and the bitstreams were provided by Nokia.

JVT-P044 [H.-C. Choi] Embedded quant. for improved SNR scalabilityAn embedded quantization was presented. The significant DCT coefficients of the residues of each SNR enhancement layer are divided into 3 groups and quantized into three values, -1, 0, and 1. In order to quantize and de-quantize the DCT significant coefficients, the proposed method calculates quantization intervals and reconstructions values in the sense of R-D optimization. Then, DCT

115

significant coefficients of each enhancement layer are quantized and de-quantized using the values obtained by the proposed method. This method computes the quantization intervals and reconstruction values in the sense of R-D optimization. Improvement of 0~1.4dB were claimed in the first SNR enhancement layer.

No decoder. Further work is encouraged.

JVT-P086 [W.-J. Han] Distribution-based inverse quantizationThe quantization and inverse quantization process defined in JSVM assumes the Laplacian distribution with fixed decoding points. In this contribution, the PDF of the coefficients at the decoder side is estimated using the histogram of the quantized levels. In addition, a simplified approximation of the coefficient histogram is proposed to minimize the complexity overhead. The reconstruction points are the derived by equally dividing the PDF areas in the given quantization interval. The PSNR gain is claimed to be up to 0.8 dB at the Qp value of 8 and 0.2 – 0.3 dB at the relatively low Qp values from 24 to 28.

Add to CE.

JVT-P095 [T. Rusert, M. Wien] Locally adaptive quantiz.: concept and first resultsA concept for bit rate allocation with locally adaptive quantization for the JSVM is presented. The error-propagation characteristics within the MCTF scheme are considered on a macroblock basis by modified computation of the distortion measure used for rate-distortion optimization. In an analytic approach, a performance gain of up to 0.3 dB is observed for four decomposition levels. In a model-based approach, a performance gain of up to 0.2 dB is observed for one decomposition level.

Further work is encouraged.

3.6.10 CE 10 & related docs: Extended spatial scalabilityJVT-P012 [S. Sun] CE10Pt1: Direct interpolation for upsamplingThis contribution is a response to CE10 part 1. It shows performance of the improved upsampling process applied to extended spatial scalability. The proposed direct interpolation method is claimed to reduce computational complexity while slightly improving quality of the up-sampled pictures by up to 0.45 dB and 0.19 dB on average comparing to the JSVM2 implementation.

Gains between 0 and 0.45 dB with average being 0.19 dB

Subjective results are requested. The JVT agreed to adop an additional option into software as at this meeting. Verify performance related to activities of resampling AHG. Do not automatically move to WD unless verified at the next meeting.

JVT-P020 [J. Vieron] CE10Pr1: Cross-verif of JVT-P012 direct upsampReports the cross-check result for the proposal 1 of CE10 described in document JVT-P012, related to direct texture upsampling method.

JVT-P018 [E. Francois, J. Vieron, S. Sun] CE10: Report of resultsThis document reports the results of CE10 related to Extended Spatial Scalability. The goal of this CE is to improve the upsampling processes applied in extended spatial scalability as described in the JSVM2. This basically addresses the intra texture, residual and motion upsampling.

116

JVT-P019 [J. Vieron, E. Francois] CE10: Unified motion upsampling in extended spat scalThis document relates to CE10, dealing with Extended Spatial Scalability. It corresponds to proposal 2, focused on motion upsampling, in case of cropping and non dyadic spatial scalability. The proposed solution aims at unifying the two tools of Extended Spatial Scalability, that is, spatial scalability with ratio 3/2 and MB aligned cropping, and spatial scalability with any ratio and any cropping.

The JVT agreed to adopt.

JVT-P074 [S.-W. Park, J.-H. Park, D.-H. Yoon, B.-M. Jeon] CE10: Verif. of Thomson proposal JVT-O041/O042The main purpose of this document is to verify the performance of the CE 10 related to improving upsampling process applied in Extended Spatial Scalability (ESS) as described in the JSVM2. In this document, we focus on verifying the results of Thomson’s proposal (JVT-O041/O042) of CE 10. The proposal consists in providing a complexity reduction by fully unifying ESS_3_2 with a limited enhancement / base layer ratio of 3/2 and ESS_generic supporting any ratio. The three ratios of 3/4, 2/3 and 3/5 are considered and generic solution and unified solution are compared for each of those ratios.

Continue CE.

3.6.11 CE 11 & related docs: FGS orderingJVT-P054-LL [J. Ridge, M. Karczewicz, Y. Bao, X. Wang] CE11: Core experiment results for FGS reordering

There were reportedly two participants in CE11: HHI and Nokia. No results or technical inputs were provided. In the case of Nokia’s previous proposal, coding the FGS layer by macroblock rather than by 4x4 blocks appeared to negate much of the previously observed gains. Further study within the new structure is required before the matter is again considered by the JVT.

JVT-P090 [B.-K. Lee] Adaptive FGS processUtilizes deblocking filter for Intra_BL prediction.

Shows bitrate savings of 6-7% at 1st FGS enpoint.Add to CE on inter-layer texture prediction.

Stop CE 11.

3.6.12 CE 12 & related docs: Weighted prediction FGS coding of closed-loop P framesJVT-P055-L [Y. Bao, M. Karczewicz, X. Wang, J. Ridge] CE12: Weighted pred. FGS coding closed-loop P framesDue to later release of the JSVM2.0 software, Nokia reported that it was not able to finish this core experiment in the new software on time. Improving the FGS coding performance of SVC under low-delay constraints was reported to be very critical for many applications. Nokia also reported seeing more interests and new contributions in this direction from other companies. Nokia requested continuation of this core experiment

Continue CE12 on low-delay SNR scalability.

JVT-P046-L [J.-Z. Xu] Alternative to FGS quant. method of JVT-O054.

117

An alternative FGS coding method to JVT-O054 is presented. It is claimed that it can improve the coding performance compared to the proposed one in JVT-O054 and meanwhile effectively reduce the decoding complexity of motion compensation operation for enhancement macro-block. It is implemented based on Progressive FGS (PFGS) with macro-block based drifting control. Single-loop decoding is claimed.

Add to CE12.

JVT-P087 [W.-J. Han] PFGS with bilinear interpolation filterCompared to the current implementation of FGS technique in JSVM, progressive fine grain scalability (PFGS) uses the FGS layer as the reference frames to improve the coding efficiency. However, the computational complexity is the major burden of PFGS since motion compensation is needed for each FGS layer. In this contribution, the scheme for estimating the FGS residual in the PFGS scheme is modified. In addition, a less complex bi-linear interpolation filter in the FGS layer is used to reduce the overall complexity of PFGS-based schemes. From the experimental results, it can be shown that the use of a bi-linear interpolation filter does not degrade the PSNR values in the FGS layer, and the modified residual estimation process provides up to 7% bit-saving compared to the direct PFGS residual computation process.

Add to CE12.

3.6.13 High-Level SyntaxJVT-P031 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12306) Syntax for FGS pass fractioningA syntax modification for allowing FGS pass fractioning is proposed. With that, an FGS enhancement pass may be split over several NAL units. This syntax modification is claimed to be justified for enabling dead sub-streams implementation and to help proposing Medium Grain Scalability at the NAL unit level.

Presentation has been made that showed a consensus of the break-out group containing a design similar to JVT-P031 but introducing a new slice_type at the benefit of saving 1 bit. JVT experts comment was to spend the bit and merge the slice_type into one slice_type, i.e., amending the PR slice_type.

The use of base_id_plus1 was adjusted to account for FGS fragments.

The JVT agreed to adopt as described in JVT-P112.

JVT-P112 [S. Pateux, Y.-K. Wang, P. Amon] Agreed Syntax for FGS pass fractioning.Document drafted to capture syntax from side activity.

JVT-P032 [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] (M12307) On high-level syntax for SVCSome discussions with examples and syntax improvements for the high level syntax of NAL units were proposed. Based on these discussions it was proposed to always use the extended NAL unit header (i.e. having simple_priority_id, dependency_id, temporal_level and quality_level) and thus removing the extension_flag introduced in the NAL unit header. Further the mapping between simple_priority_id and (dependency_id, temporal_level, quality_level) should be relaxed to get more flexibility.

118

JVT-P100 [P. Amon, A. Hutter, T. Rathgen] High-level syntax for SVCThree different tools for the high-level syntax of SVC are proposed. The first topic addresses the scalability description in the NAL unit header. First, the current status of the discussion is briefly reviewed, which is a variable length design. Then two alternate solutions are presented, which basically introduce more flexibility in the usage of the NAL unit header. The second tool describes the concept of Super-NAL units. These container elements include all NAL units of one scalability level (the same spatial, temporal, and SNR ID). This design facilitates scalability operations and supports transport. The third tool deals with the design of FGS NAL units to allow medium and fine grain SNR scalability with the high-level syntax.

a) Introduction of FGS fragment NAL units. The JVT agreed to adopt this.

b) NAL unit header

a. 1 Byte onlyb. 2 Bytes onlyc. 1 or 2 Bytes

Router/proxy: drop the right packetsSink: drop the right packetsSource: thinning

Problem: Can we expect routers/proxies to be capable ofa) do more than dropping packets that are larger than a given value (1 Byte) – fixed pathb) parse an SPS to derive alternative adaptation paths (1 Byte) – variable path at cost of parsing

SPS and established 1D-3D relationshipc) use the explicit 3D scalability information to determine the adaptation path (2 Byte) –

variable path at the cost of 1 Byte overhead per NAL unit

[ EXT DS P5 P4 P3 P2 P1 P0 ][ EXT DS Q5 Q4 Q3 Q2 Q1 Q0 D2 D1 D0 T2 T1 T0 L1 L0 ]

JVT-agreed decision: keep switchable version of NALU ext header with different meaning of P and Q.

JVT agreed: Removal of "dead substream" SEI message.

JVT-P061 [Y.-K. Wang, M. M. Hannuksela] Signaling of scalab. info.A mechanism for signaling of scalability information is proposed, targeting for easy and convenient system-level operations on scalable video streams. The concrete proposals enabling the signaling mechanism are 1) An improved syntax for the scalability information SEI message specified in the JSVM 2 (JVT-O202), 2) A new SEI message called sub-picture scalable layer information SEI message, 3) A restriction on the use of the three sub-sequence related SEI messages.

Discuss in connection with JVT-P042. The JVT agreed to adopt as described in JVT-P113.

JVT-P113 [Y.-K. Wang, T. C. Thang] Agreed Signaling of scalab. info. SEI messageThis document presents a revised signaling mechanism of scalalability information, which includes a revision to the syntax of the scalability information SEI message presented in JVT-P061 by addressing identification of multiple interested regions presented in JVT-P042.

119

JVT-P062 [Y.-K. Wang, Y.-L. Bao] Indication of non-required pics.A new SEI message is proposed for indication of non-required pictures before their decoding, such that the decoder can avoid decoding and buffering of the non-required picture while at the same time low delay is allowed.

Relationship with fragmented NAL units and dead substreams has been clarified and SEI message has been adjusted accordingly. The JVT agreed to adop adjusted SEI message as decribed in JVT-P062r1. JVT-P062r1 has been presented.

JVT-P079 [D.H. Yoon, B.-Y. Jeon, J.-H. Park, S.-W. Park] Key picture indication in AVC compatible base layerThe JSVM2.0 key pictures use closed-loop motion compensation. The reference pictures for the key picture are the base representation without FGS enhancement. The reference pictures of non-key picture are the base representation + FGS enhancement. It is claimed that the way how to indicate key picture for slices conforming to AVC ( the base layer of SVC ) is not clear so far, since key_picture_flag is defined only in slice_header_in_scalable_extension. A solution using nal_ref_idc is proposed.

Adopted by the JVT.

JVT-P080 [S. Jeong, M. Park, G. Park, K. Kim] Flag bit in SEI for adaptive GOP structureAdaptive GOP Structure was adopted into JSVM as non-normative because it dose not require any modification for the decoder. It is claimed that it requires a modification on the extractor. In order to use Adaptive GOP structure in JSVM, the information whether Adaptive GOP Structure is used or not is required. Therefore, a flag bit in Scalable Information is proposed in order to notify whether Adaptive GOP Structure is used or not to the extractor. This flag bit is added into Scalable Information SEI, because it used only at the extractor.

Two methods proposed:a) reversing the numberingb) flag to indicate lowest resolution – added syntax to repair old concept (GOP structure)

representing syntaxConsult with the editors. Clean syntax to accommodate AGS. Adopted by the JVT as described in JVT-P114.

JVT-P114 [J. Reichel, S. Jeong] Simplified syntax for AGSIn this document was reportedly shown that the current syntax allows for AGS, but that some basic modification of the syntax could improve the flexibility and the efficiency of the JSVM.

3.6.14 VLC-based entropy codingJVT-P056-L [J. Ridge, M. Karczewicz, Y. Bao, X. Wang] Variable length codes for SVCAs noted at the previous JVT meeting, context-adaptive variable-length codes (CAVLC) is a feature currently missing from the JSVM software. CAVLC is incorporated into the JSVM. Base-layer compatibility is verified with H.264/AVC by decoding test sequences using H.264/AVC JM 9.6. For enhancement layers, the performance is claimed to be in line with the single-layer case.

2 parts:a) CAVLC for base layer. Implementation of existing tools. Adopted by the JVT into JSVM

software.

120

b) New tools for scalable coding. Create CE on VLC-based entropy coding to verify the results. Add tool to the software at this meeting to enable incorporation into other proposals. Test conditions to be provided by Nokia at the same bitrates as for the CABAC case.

Consider implications of mixing CAVLC and CABAC into one SVC bitstream.

3.6.15 Weighted predictionJVT-P064 [P. Yin, J. Boyce, P. Pandit] Weighted prediction for SVCJSVM2.0 uses the same pred_weight_table() syntax in the scalable extension as in AVC, which indicates that the enhancement layer can have a different pred_weight_table() from its previous layer. In this contribution, it is proposed to re-use the base layer weighting parameters for enhancement layer if the corresponding base layer exists, i.e., to remap the pred_weight_table() from the base (or previous) layer to pred_weight_table() in the current enhancement layer. It has the advantage of saving memory, complexity and can achieve a slight coding efficiency improvement at very low bit rates.

JVT-P076 [K.-H. Lee] Multi-layer weighted predictionMCTF is a temporal wavelet process with motion compensation. MCTF can be used in multi layer structure scalable video coding. It is proposed that weight for motion prediction in upper layer can be estimated from base layer information.

Same as JVT-P064.

2 parts:a) Weighted prediction for the base layer. Adopted by the JVT into JSVM software.b) New tools for scalable coding. Adopted by the JVT.

JVT-P071 [A. Segall] Verif. of JVT-P064 weighted pred.Verification of JVT-P064. Bit-streams and a binary decoder were provided. It has been verified that the bit-streams are decodable and bit rates and PSNR values have been measured. The bit rate and PSNR information is enclosed in a supporting document.

JVT-P065 [P. Yin, J. Boyce, P. Pandit] Weighted prediction for SVC MCTF update stepThe update step is investigated when weighted prediction is used in prediction step. First, the weights for the MCTF update step are theoretically derived based on the weights from prediction step. Then simulations are conducted to study two issues using fading sequences. The first issue is if the MCTF update step can improve coding efficiency compared to MCTF without update step when weighted prediction is used in prediction step. The experiment shows that MCTF update can achieve only a little gain compared to MCTF without update for fading sequences when weighted prediction is used in prediction step. The bigger gain can only be achieved by the sequences of longer fades which mainly contain global motion with larger GOP Size. The second issue is if the update step should incorporate the weighting parameters used in the prediction step since JSVM2.0 applies adaptive update weights to control the update strength. The experiment shows that MCTF update with derived weights from prediction step does not achieve any gain compared to MCTF update without derived weights. But since MCTF update with derived weights from prediction step involves more computational complexity, it is simpler to perform MCTF update without derived weights.

Do not apply the weights for the update step. Adopted by the JVT.

121

3.6.16 Other subjects

JVT-P045 [I.-H. Shin, H.-W. Park] Adapt. upsamp. for spat. scal.A new up-sampling method for SVC is proposed. A type-1 DCT up-sampling method is employed. An adaptive filtering method of luminance component in the type-1 DCT up-sampling is used, which applies different weighting parameters to DCT coefficients. The adaptation is claimed not to incur an overhead in computational complexity at the decoder side. The weighting parameters of the up-sampling matrix are transmitted as side information.

Only results for intra-frame coding. The JVT agreed to add to CE on inter-layer texture prediction.

JVT-P051 [J. Jia] Imp. coding eff. by polyphase downsamplingA method for improving SVC coding efficiency on the basis of polyphase down sampling (PD) is described. The presented method proposes to implement PD on the residual signal before integer transformation in the SVC enhancement layer. Experiments based on JSVM2.0 are performed with different test sequences on Palma test points. It is claimed that the presented method gives 0.08 – 0.2dB improvement over the tested set of sequences when comparing at the same bit rate. It also claimed that the presented method effectively improves the coding efficiency of FGS enhancement layer by 14% average bit rate reduction at a 0.07dB PSNR loss in enhancement layer.

The JVT agreed to add to CE on inter-layer texture prediction.

3.7 JVT Non-SVC normative modifications3.7.1 Improved 4:4:4 codingJVT-P016 [Y.-L. Lee] Lossless intra coding for improved 4:4:4A new lossless Intra coding method based on DPCM (Differential Pulse Code Modulation) is proposed. The High 4:4:4 profile uses a multi-directional spatial prediction method to reduce the spatial redundancy by using block boundary pixels. In the proposed lossless Intra coding method, the spatial prediction residuals are transformed to the residual DPCMs. Also, the proposed method based on DPCM is designed not to break the decoder pipeline processing. The proposed lossless Intra coding based on DPCM is applied after the existing H.264 horizontal and vertical prediction. The proposed lossless Intra coding method based on residual DPCM reduced the bit rate by approximately 12.53 % in comparison with the lossless Intra coding method included in the H.264 standard.

Method of improving lossless intra coding. Based on concepts presented at Redmond meeting JVT-L017. Applied to block sizes used in FRExt (4x4, 8x8, and 16x16). Changes operation in two of the intra pred modes. Draft text provided – about one page of text.

Performance – 12.5% savings on average for intra. Pretty good. Any complexity problem? Doesn't seem like much complexity. On the surface, seems like little added complexity. Some misgivings expressed (the need to use mode-customized transforms in two cases and not in others).

If we had a new profile, would we want this to replace the current intra in it? Not sure yet.

Proponent says when testing on 8-bit YUV4:2:0, this method performs roughly same (maybe a little better) than JPEG-2K lossless.

Comment: Have you tried to run it on YCoCg? No.Comment: Should test more thoroughly.

122

Comment: Current intra design may not be what we would want for other reasons for 4:4:4 – see other contribs.

Generally favorably disposed, but needs further study and depends on other 4:4:4 plans.

Later verbal report – when using this method with new intra proposed coding methods (e.g., more pred modes available for chroma), seems to work just as well in that context and maybe even better. Should study more.

JVT-P017-L [H. Yu, L. Liu] Advanced 4:4:4 profileThe Advanced 4:4:4 Profile proposed in JVT-O013 has been implemented in the JM software. The presented test results are used to claim that the proposed Advanced 4:4:4 Profile outperforms the current High 4:4:4 Profile. The latest development of the proposed Advanced 4:4:4 Profile are discussed and new test results are presented.

Proposes not using RCT.

Two options proposed for intra – same intra spatial prediction modes available for all 3 channels, with all 3 channels handled like luma is now, motion comp interpolation for chroma uses same method as for luma, separate CABAC context models for each color component, identical to current luma operation, same luma and chroma resid syntax.

"Solution 1": Intra predictors selected separately for each color component."Solution 2": Intra predictor type selected once, and then same mode applied for each color component.

Summary of relative perf: Solution 1 beats Solution 2 in quality by average 0.2 dB at 45 dB and above (which is not a very impressive amount).

Asserted that, in applications of interest to proponent (very high quality, high resolution), 45 dB is realistic (perhaps even a minimum). Average intra-only improvement roughly 1.5 dB (e.g., 25% bit rate savings) in that neighborhood, depending on what is compared. Most of that is retained down to 40 dB or less.

Similar efficiency differences with inter prediction.

Artifacts noted in some sequences.

RCT results seem rather negative and mystifying.

Not cross-verified.

JVT-P048-L [W.-S. Kim+] Further experimental results on 4:4:4 codingPerformance of coding tools using various images of different characteristics – film grain noise, thermal noise, sampling, and cropping, etc is discussed. After brief description about RCT coding method, additional experimental results on the 4:4:4 chroma format materials are shown aiming to verify the efficiency of the RCT coding over various kinds of image sequences.

JVT-P049-L [W.-S. Kim+] Study on advanced 4:4:4 coding

123

Coding methods such as inter-plane prediction and single mode prediction are proposed aiming at improving the coding efficiency of the current FRExt 4:4:4 video coding technology, especially for RGB coding using RCT. A new work item of advanced 4:4:4 video coding is suggested in addition to the current standard.

JVT-P083 [Y. Yamada, S. Sekiguchi, Y. Isu, K. Asai, T. Murakami] Consideration on intra coding eff. of High 4:4:4 pr. (not Sunday afternoon)Discussion of intra coding efficiency of the High 4:4:4 profile. At the Busan meeting, there was a discussion on the need of advanced 4:4:4 profile due to a loss of coding efficiency at very high bitrate. There was a discussion focus on intra coding efficiency of the High 4:4:4 profile especially when the input source signal is not YUV basis. For the applications targeting very high quality and resolution, color conversion to YUV space may be a limitation for further coding gain due to an information loss derived from color matrix conversion. The intra coding performance of the High 4:4:4 profile for XYZ formatted content was tested, which is adopted by Digital Cinema Initiative (DCI) as the signal representation for digital cinema contents. It is claimed that the intra coding performance for 4:4:4 high quality contents could further be improved by careful design of the intra prediction scheme suitable for 4:4:4 contents.

JVT-P088 [T. Wedi, S. Kondo, H. Saito] 4:4:4 Intra-only codingHigh 4:4:4 profile uses the same mode for all three color-components of a MB [not true]. Thus, the color components cannot be coded independently and the possibility of parallel processing is limited. This limitation is claimed to be a drawback, because parallelization is essential, especially for future professional cameras that record video content at incessant higher resolutions. An Intra-only method is proposed that allows independently encoding and decoding of each color component. Therefore, separate mode information is associated to each color component on MB basis. In order to restrict the memory consumption, it is further proposed to keep the interleaving of color components on MB-level. Thus, the possibility of parallel processing for Intra only applications is significantly enhanced. It is claimed that the maximum (worst-case) overhead due to assignment of separate modes for each color component is negligible for professional applications.

JVT-P102-L [H. Yu] Artifacts seen in some Viper 4:4:4 sequencesSome severe artifacts in some Viper 4:4:4 JVT test sequences are reported. These sequences are “breeze” and “kungfu”. These problems are discussed and some preliminary results from the investigation are presented.

Recommended not to use these parts of the Viper material. Look for different scenes to replace those two.

Discussion:Create 1 or 2 new profiles. 1. Intra-only2. Inter coding added to first profileTimeline: CD in October assessed to be possible and desirable.

Idea to have a meeting of the AHG for subjective tests in Indianapolis at Thomson facilities. Involvement of MPEG test group suggested. Further discussion needed.

The JVT agreed to create a JVT resolution on consideration of removal of the current Hi444 Profile

124

3.8 JVT Non-SVC non-normative aspects3.8.1 Encoder optimization techniquesJVT-P014 [H. Schwarz, D. Marpe, T. Wiegand] Hierarchical B PicturesAn investigation of H.264/MPEG4-AVC compatible encoding with hierarchical B pictures is presented. The memory requirements are analyzed, the modifications of the encoder control compared to the H.264/MPEG4-AVC Test Model are described which improve the coding efficiency for hierarchical B pictures, and present a comparison of the coding efficiency for different GOP sizes (number of hierarchy stages) and a wide range of test sequences. It is claimed through simulation results that in comparison to the classical “IBBP…” coding, the coding efficiency can be improved by up to 1.5 dB just by modifying the coding structure. It is also claimed that the usage of hierarchical B pictures does not only improve the objective quality for the majority of test sequences, but also the subjective quality especially for sequences with fine-detailed slow/regular moving image regions.

Using their own software.

Remark: IBBP is dead. Long live hierarchical B pictures.

Remark: JVT members should take note, and point out the deficiencies of any comparisons of other codecs to AVC that do not account for the ability to use hierarchical B pictures in AVC.

To achieve temporal scalability and use the minimum amount of reference picture memory, can use long-term reference pictures. Other methods might be possible.

f = 1/3 in Key pictures, 1/6 in non-key. QP = QP of lower layer + 1 (except +4 for 1st enhancement level).

Sometimes gains in perceptual quality even when not in PSNR.

JVT-P015 [G. J. Sullivan, J. Lou, M.-T. Sun] Adaptive quantization yielding 10 dB gainResults obtained from experimentation using adaptive quantization in video encoding are described. The results of a small number of tests are reported on the effect of adaptive quantization on luma and chroma fidelity with different adaptive-update weighting factors. The amount of luma-component quality improvement that is obtained by use of this technique is claimed to be about 11 dB (10 dB at equivalent bit rates). This substantially exceeds the 3.7 dB anticipated improvement as calculated using high-resolution quantization theory analysis described in the contribution. The chroma fidelity difference for use of the technique is claimed to be in the 5 to 6 dB range. Results are reported for one particular video sequence – however, one additional sequence is reported to have been checked, and it is asserted that the results are likely to be roughly similar for others. The results are reported to have been approximately verified independently by Shijun Sun of Sharp. Remarks on intra and lossless coding are also provided.

JVT-P021 [X. Yi, J. Zhang, N. Ling, W. Shang] Improved & simplified fast ME for JMIn order to reduce encoding time, the current JVT reference software JM adopted a fast motion estimation method called UMHexagonS. In this document, an improved and simplified fast motion estimation scheme is proposed that claims to further speed up the encoding process and enhance the rate-distortion performance. Simulation results are shown to underline the claim that with similar or even better rate-distortion performance, the proposed method reduces motion estimation time by about 55% and 94% on average when compared with those of UMHexagonS and Fast Full Search

125

respectively. In addition, a bit rate reduction up to 18% is claimed when comparing with that of full search in low complexity mode.

Based on existing method proposed by Tsinghua Univ, but simplifies search flow and provides early termination conditions. Not clear whether PSNR and bit rate comparisons are using BD method – results may actually be better than reported. Software provided in JM 9.6 (changes in 3 files), lines of code reduced.

Seems like good work. Reportedly partly verified by Tsinghua. Try to integrate into ref soft hhin addition to current method.

JVT-P026 [A. M. Tourapis, H.-Y. Cheong, P. Topiwala] Fast ME in the JM reference softwareThe scheme presented in JVT-E023, with some minor enhancements is proposed.

Integrated reference software has been provided, although not yet fully studied by interested parties. Having both versions in the software for a while seems advisable to give the opportunity to study the relative capabilities of the two methods. Let's keep two methods in the software and study their relative benefits. Perhaps eventually we will remove one of the two methods. Coding efficiency performance reported to be roughly the same as existing Tsinghua method. May be faster than the JVT-P021.

Keep all three fast methods in the reference software for a while to allow opportunity for study and comparison – hope to eventually just have one method. For now, we'll have both the JVT-P021 and JVT-P026 methods and the older method integrated with config file switching.

The group discussed creating an AHG to conduct futher study, but decided not to pursue this idea.

JVT-P111 [G. Sullivan] Decoder Inference of Optimal Reconstruction Values for DZ+UTQ Quantization of Laplacian Source Random VariablesThis document, provided in response to an inquiry from the chair during the Poznan JVT meeting, shows (using one page of text with 12-point font) that it is possible for a decoder, using only quantized histogram statistics and knowledge of the quantizer threshold values, to infer MSE-optimal quantization reconstruction offsets for a typical (Laplacian) pdf source model. The results can reportedly be used in an FGS-embedded context or a conventional non-embedded context.

Information document only. No action needed.

3.8.2 Improved Error resilienceJVT-P023 [V. Bottreau] CABAC variant and options for bit error resilienceCABAC variants are proposed as options to support error-resilient decoding. It also presents a CABAC decoding algorithm making use of these options. The decoding approach is based on a Maximum-A-Posteriori (MAP) estimation algorithm. The trade-off between compression efficiency is controlled in a straightforward manner with a single parameter. When this parameter is set to zero, the performance is identical to the current CABAC implementations. Similarly, the trade-off between error-resilience and complexity is controlled by another parameter exploited by pruning techniques in the decoder. The performances of the decoding algorithm in the coder/decoder JM software are assessed with different error detection method against those obtained with classical CABAC hard decoder.

126

Comparison to better anchors needed that use more already standardized error resilience features. Common test conditions should be applied. Further work is encouraged.

JVT-P072 [S. Bandyopadhyay, Z. Wu, P. Pandit, J. Boyce] Frame loss concealmentThe current JM 9.6 software has limited error concealment support. In particular, the reference decoder software cannot handle entire frame (or picture) losses. Two error concealment algorithms are proposed to be be added to the JM to handle frame loss – frame copy and motion vector copy. The proposed error concealment algorithms were tested using the packet loss model of VCEG Q15-I-16r1. The motion copy method reportedly outperforms the frame copy method by up to 4.6 dB.

Great to have this work – encourage integration – adopt into reference software & JM.

JVT-P050 [J. Jia] Polyphase downsamp.-based redundant slicesA polyphase down sampling (PD) based redundant slice coding method for SVC error resilience is proposed. The proposal is to implement PD in SVC redundant slice coding process, then encode and send a part of the DCT coefficients. In decoding process, the uncoded pixels in redundant slice are reconstructed by both motion compensation and spatial interpolation. Experiments based on JM9.6 are performed on different test sequences in error prone environment with average packet loss rate of 3%, 5%, 10% and 20%. It is claimed that the PD-based redundant slice coding method gives 1dB improvement on average PSNR and 0.5-5.2 dB improvement on frame by frame PSNR over the conventional redundant slice coding method based on the same bit rate.

Comparison to better anchors needed that use more already standardized error resilience features. Common test conditions should be applied. Further work is encouraged.

3.8.3 VUIJVT-P103 [A. Rodriguez] Sample aspect ratio supportThree aspect_ratio_idc values are proposed for picture resolutions used in broadcast applications and HD DVD.

Seems like a good idea to include these in the next available amendment. Agreed by the JVT.

3.9 Withdrawn JVT documentsThe following document registrations were withdrawn: JVT-P027, JVT-P033, JVT-P039, JVT-P058, JVT-P073, JVT-P099, JVT-P101, and JVT-P110.

3.10 Updating of JVT group documentsJVT-P024 [T. Suzuki] Editor's proposed draft FRExt conformanceJVT-P025 [T. Suzuki] Prop list of changes of AVC conformanceThe conformance specification, including FRExt conformance, were consented in ITU-T and it is under trial publication. However, some minor bugs were reported after ITU-T’s publication. All problems and its fixes are summarized in the provided http file. CRC32 and time stamp is attached for each bitstream to identify the latest version. All those fixes have already been incorporated into the ISO version of the conformance specification during editing period before the ISO publication.

The JVT agreed to remove cpb_removal_delay from paragraph in clause 6.5.6 as proposed in JVT-P025.

3.11 JVT operating rulesThe following clarifications/adjustments of JVT operating rules have been approved by the JVT.

127

All submissions must be made in JVT-Qnnn.zip format with the word docs, excel sheets and other information being in the zip container. A revision of the doc shall be made by adding the revised doc to the zip container. (The document must also contain an abstract and be accompanied with an e-mail notification containing title, authors and abstract (identical to the one in the doc) which is no longer than 200 words written in 3rd person.)

Independent verification (necessary for adoption of a proposal) is provided either through a) independent implementation by 1 or more company different than the proponent based on

the textual description (after adoption, both decoder source code versions must be made publicly available and one encoder version)

b) providing source code to all CE participants prior to the meeting (CEs can only be joined at the meeting, when the CE is created. CEs are created at each meeting and last until the next meeting.)

For every SEI message and every syntax element that are currently in the SVC draft, a show case has to be provided in order to retain it in the JSVM/WD. If such a showcase is not provided at the next meeting for an SEI message or parts of it, the SEI message or the respective parts will be removed from the JSVM/WD.

A first CE description must be available at the last day of the meeting. Changes of the CE description are only allowed until 1 month prior to the next meeting. These changes must be of evolutionary characteristic relative to the input documents on which the CE is based and must be agreed by those who contributed the respective input document(s) or be added as an option.3.12 List of AdoptionsPerson listed in bracket is responsible for provisioning of text and software integration.

3.12.1 AHG report related adoptions JVT-P006 AHG Report: JSVM & WD text. Approval of JSVM 2 text and WD 3 text. New

procedure: Use Annex S in JSVM for all normative elements. This can be copied into the WD right away.

JVT-P005 AHG Report: JSVM s/W and new func. integ. JSVM 2.1 software has been approved.

JVT-P008 Err resil test cond's & apps. Testing conditions have been adopted as described in JVT-P206 subject to further refinement until 45 days prior to the JVT Nice meeting.

JVT-P009 Testing conditions for future benchmarking of SVC development and core experiments as described in JVT-P205.

3.12.2 Normative SVC adoptions [J. Reichel] JVT-P028 CE01: Improved residual picture buffer management / JVT-P040 [M.-H.

Lee] Progressive MCTF [M. H. Lee] JVT-P037 CE01: Reduced Memory MCTF: clipping of the reference signal prior to

update to be represented by 8 bits per sample. [S. Sun] JVT-P013r1 CE02Pt2: Deblocking filter for I_BL Spatial SVC [Y. Bao] JVT-P089 CE03: Report on symbol pred. of resid. pred. flag: Adopt, by adding a

context model that is conditioned on the CBP value of the base layer. [X. Wang] JVT-P047 CE06: Proposed approach to CE06 / JVT-P052 CE06: Simplified update

step operation for MCTF: Adopt unified solution (JVT-P052 and JVT-P047) as described in JVT-P052r1.

128

[D. H. Yoon] JVT-P078 Clarification of intra_base_flag: In single-loop decoding, when Intra_BL prediction is not possible, the flag that indicates the use of Intra_BL prediction should not be present.

[E. Francois] JVT-P019 CE10: Unified motion upsampling in extended spat scal [S. Kervadec] JVT-P031 (M12306) Syntax for FGS pass fractioning / JVT-P100 High-level

syntax for SVC : Adopt as described in JVT-P112, Introduction of FGS fragment NAL units. Adopt. , Decision: keep switchable version of NALU ext header with different meaning of P and Q. Removal of "dead substream" SEI message.

[Y.-K. Wang] JVT-P061 Signaling of scalab. info. / JVT-P042 CE08: Spatial scalability of multiple ROIs. Adopt as described in JVT-P113.

[Y.-K. Wang] JVT-P062 Indication of non-required pics.: Adopt adjusted SEI message as decribed in JVT-P062r1.

[D. H. Yoon] JVT-P079 Key picture indication in AVC compatible base layer [S. Jeong] JVT-P080 Flag bit in SEI for adaptive GOP structure: Adopt as described in JVT-

P114. [J. Boyce] JVT-P064 Weighted prediction for SVC / JVT-P076 Multi-layer weighted prediction:

2 parts: a) Weighted prediction for the base layer. Adopt into JSVM software, b) New tools for scalable coding. Adopt.

[J. Boyce] JVT-P065 Weighted prediction for SVC MCTF update step: Do not apply the weights for the update step. Adopted.

3.12.3 Non-Normative SVC adoptions [H. Schwarz] JVT-P059 Comparison of MCTF and closed-loop hierarch. B pics., non-normative [N. Cammas] JVT-P029 (M12304) CE05: CE report on quality layers: Non-normative part

adopted.

3.12.4 Software adoptions that require further verification (do not automatically move to WD unless verified at the next meeting)

[S. Sun] JVT-P012 CE10Pt1: Direct interpolation for upsampling: Adopt an additional option into software as at this meeting. Verify performance related to activities of resampling AHG.

[J. Ridge] JVT-P056-L Variable length codes for SVC: CAVLC for base layer. Implementation of existing tools. Adopt into JSVM software, New tools for scalable coding.

3.12.5 Normative non SVC-related adoptions [T. Suzuki] Adopt to remove cpb_removal_delay from paragraph in clause 6.5.6 as proposed in

JVT-P025 [T. Suzuki] Approval of new conformance bitstreams are reported in JVT-P003. (Bitstreams

with hierachical GOP structure)

3.12.6 Non-Normative non SVC-related adoptions [X. Yi] JVT-P021 Improved & simplified fast ME for JM [A. Tourapis] JVT-P026 Fast ME in the JM reference software [J. Boyce] JVT-P072 Frame loss concealment

3.13 List of JVT AHGs establishedAHG on Project management and errata [Chairs: G. J. Sullivan, J.-R. Ohm, A. Luthra, T. Wiegand]AHG on JM text and reference software [Chairs: T. Wiegand, K. Suehring, A. Tourapis, K. P. Lim]AHG on Bitstreams & conformance [Chairs: T. Suzuki, L. Winger]AHG on JSVM S/W and new functionality integration [Chair: G. Cook]AHG on JSVM & WD text [Chairs: J. Reichel, H. Schwarz, M. Wien]

129

AHG on Spatial scalability resampling [Chair: G. J. Sullivan]AHG on Err resilience test conditions & applications [Chair: Y.-K. Wang]AHG on Coding eff & JSVM coding efficiency testing conditions [Chairs: M. Wien, H. Schwarz]AHG on Study of 4:4:4 functionality [Chair: T. Suzuki]AHG on enhancement slice complexity reduction [Chair: L. Xiong]

Mandates for AHGs are to continue efforts of similar AHGs in previous interim period.

3.14 JVT SVC ScheduleThe original schedule for the SVC project was: WD in July 2004 CD in October 2005 FCD in March 2006 FDIS in July 2006

As the WD has been shifted by 6 months (and for various other reasons) a new schedule is proposed: WD in January 2005 (has already happened) CD in April 2006 FCD in July 2006 FDIS in January 2007The new schedule is approved by the JVT.

3.15 Key JVT resolutions of the meetingThe JVT nominates the following persons as editors for H.264 Annex G | 14496-10:200x/AMD1: Julien Reichel, Heiko Schwarz, Gary Sullivan, Thomas Wiegand, Mathias Wien.The JVT and changes the work plan for H.264 Annex G | 14496-10:200x/AMD1 as follows: ITU: AAP in January 2007 time frame; MPEG: PDAM – April 2006; FPDAM – July 2006; FDAM – January 2007.The JVT has concluded that the creation of a new 4:4:4 profile of AVC having substantial improvement in compression efficiency for high-fidelity coding is feasible and may be justified. Further information is provided in document JVT-P204. This conclusion may call into question the appropriateness of the current AVC High 4:4:4 Profile specification. National bodies are requested to comment on the possibility of rectifying the situation by a corrigendum that would replace the current High 4:4:4 Profile according to the new design information, or by an amendment that would add one or more additional profiles.

3.16 JVT-related liaison activity by WG11 parent bodyIt is noted that the parent body under whose auspices the meeting was held sent two liaison letters on behalf of the JVT.

1. One to 3GPP SA 4 on experiment conditions for corrupting/erasure channel testing2. One (formatted as five) to DVB TM-AVC, ATSC S6, SCTE DVS, DVD Forum WG-1, and

Blu-ray Disc Association TEG-2 on the aspect ratio indications (see discussion of JVT-P103).

3.17 JVT SVC SW integration planStart: JSVM 2.1 software available (probably 05/08/05) [H. Schwarz] JVT-P059 Comparison of MCTF and closed-loop hierarch. B pics., non-normative2 days

130

[J. Reichel] JVT-P028 CE01: Improved residual picture buffer management / JVT-P040 [M.-H. Lee] Progressive MCTF

[J. Reichel] JVT-P080 Flag bit in SEI for adaptive GOP structure: Adopt as described in JVT-P114.

3 days

[X. Wang] JVT-P047 CE06: Proposed approach to CE06 / JVT-P052 CE06: Simplified update step operation for MCTF: Adopt unified solution (JVT-P052 and JVT-P047) as described in JVT-P052r1.

[Y. Bao] JVT-P089 CE03: Report on symbol pred. of resid. pred. flag: Adopt, by adding a context model that is conditioned on the CBP value of the base layer.

5 days

[M. H. Lee] JVT-P037 CE01: Reduced Memory MCTF: clipping of the reference signal prior to update to be represented by 8 bits per sample.

2 days

[S. Sun] JVT-P013r1 CE02Pt2: Deblocking filter for I_BL Spatial SVC [S. Sun] JVT-P012 CE10Pt1: Direct interpolation for upsampling: Adopt an additional option

into software as at this meeting. Verify performance related to activities of resampling AHG. 5 days

[D. H. Yoon] JVT-P078 Clarification of intra_base_flag: In single-loop decoding, when Intra_BL prediction is not possible, the flag that indicates the use of Intra_BL prediction should not be present.

[D. H. Yoon] JVT-P079 Key picture indication in AVC compatible base layer2 days

JSVM software used for proposals on coding efficiency and error resilience using common conditions

[J. Ridge] JVT-P056-L Variable length codes for SVC: CAVLC for base layer. Implementation of existing tools. Adopt into JSVM software, New tools for scalable coding.

FMO addition [S. Kervadec] JVT-P031 (M12306) Syntax for FGS pass fractioning / JVT-P100 High-level

syntax for SVC : Adopt as described in JVT-P112, Introduction of FGS fragment NAL units. Adopt. , Decision: keep switchable version of NALU ext header with different meaning of P and Q. Removal of "dead substream" SEI message.

[Y.-K. Wang] JVT-P061 Signaling of scalab. info. / JVT-P042 CE08: Spatial scalability of multiple ROIs. Adopt as described in JVT-P113.

[Y.-K. Wang] JVT-P062 Indication of non-required pics.: Adopt adjusted SEI message as decribed in JVT-P062r1.

[J. Boyce] JVT-P064 Weighted prediction for SVC / JVT-P076 Multi-layer weighted prediction: 2 parts: a) Weighted prediction for the base layer. Adopt into JSVM software, b) New tools for scalable coding. Adopt.

[J. Boyce] JVT-P065 Weighted prediction for SVC MCTF update step: Do not apply the weights for the update step. Adopted.

Parallel track [E. Francois] JVT-P019 CE10: Unified motion upsampling in extended spat scal

131

First provided as a separate tool [05/08/05], and integrated at end of SW integration [N. Cammas] JVT-P029 (M12304) CE05: CE report on quality layers: Non-normative part

adopted.

3.18 JVT Attendance

JVT attendees (per sign-in sheet at the meeting) included the following persons (154 in total): Gary J. Sullivan (Microsoft), Thomas Wiegand (Fraunhofer HHI), Jens-Rainer Ohm (RWTH Aachen Univ.), Lee Men Huang (Panasonic), Yoshihiro Miyamoto (NEC), Jun Zhang (Santa Clara Univ.), Mike Nilsson (BT), Masato Shima (Texas Instruments Japan), Vieron Jérôme (Thomson), Edouard Francois (Thomson), Hsin-Hao Chen (OES/ITRI), Je-Woo Kim (KETI), Yong-Hwan Kim (KETI), Hyok Sung (KETI), Byeongho Choi (KETI), Diego Santa Cruz (Visiowave / G.E. Security), Francesco Ziliani (Visiowave / G.E. Security), Julien Reichel (Visiowave / G.E. Security), Shijun Sun (Sharp), Andrew Segall (Sharp), Doug Young Suh (KHU), Munchul Kim (Info. & Comm. Univ. KR), Jung Won Kang (ETRI), Zhongkang Lu (I2R), Keng Pang Lim (I2R), Leszek Cieplinski (Mitsubishi Electric), Yung-Lyul Lee (Sejong Univ.), Donggyu Sim (Kwangwoon Univ.), Woo-Shik Kim (Samsung AIT), Satoru Sakazume (Victor Company of Japan), Hiroya Nakamura (Victor Company of Japan), Jae-Gon Kim (ETRI), Mathias Wien (RWTH Aachen Univ.), Ulrich Benzler (Robert Bosch GmbH), Marco Boltz (Robert Bosch GmbH), Anthony Vetro (Mitsubishi), Emin Martinian (Mitsubishi), Ong Chin Phek (Panasonic), Dae-Yeon Kim (Sejong Univ.), Jae-Ho Hur (Sejong Univ.), Huifang Sun (Mitsubishi), Jizheng Xu (Microsoft), Per Fröjdh (Ericsson), Michael Horowitz (CoVi Technologies), Sebastien Brangoulo (ENST Paris), Jong-Lak Kim (Nextreaming), Il-Hong Shin (KAIST), Reha Civanlar (LMI), Ofer Shapiro (LMI), Alex Eleftheriadis (LMI / Columbia Univ.), Tokumichi Murakami (Mitsubishi Electric), Kohtaro Asai (Mitsubishi Electric), Viji Raveendran (Qualcomm.), Seon-Tae Kim (ETRI), Byung-Gyu Kim (ETRI), Markus Beermann (RWTH Aachen Univ.), Peter Amon (Siemens AG), Thomas Rathgen (Ilmenau Tech. Univ.), Seung-Wook Park (LG Electronics), Ji-Ho Park (LG Electronics), Doe Hyun Yoon (LG Electronics), Byeong-Moon Jeon (LG Electronics), Arnaud Bourge (Philips), Xavier Henocq (Canon), Fabrice Le Leannec (Canon), Yongju Cho (ETRI), Han Sae Song (Seoul National Univ.), Chanyul Park (Seoul National Univ.), Bae-Keun Lee (Samsung), Sangchang Cha (Samsung), Woo-jin Han (Samsung), Kyohyuk Lee (Samsung), Byeungwoo Jeon (Sung Kyun Kwan Univ.), Woong Il Choi (Sung Kyun Kwan Univ.), Steffen Wittman (Panasonic), Thomas Wedi (Panasonic), Seyoon Jeong (ETRI), Gwang Hoon Park (Kyung Hee Univ.), Ye-Kui Wang (Nokia), Min Woo Park (Kyung Hee Univ.), Kyung Il Lee (Kyung Hee Univ.), Tae Heon Bae (Info. & Comm. Univ.), Truong Cong Thang (Info. & Comm. Univ.), Justin Ridge (Nokia), Yiliang Bao (Nokia), Xianglin Wang (Nokia), K. K. Tu (ETRI), Heiko Schwarz (Fraunhofer HHI), Teruhiko Suzuki (Sony), Jan Lievens (Vrije Univ.), Stephan Pateux (France Telecom. R&D), Nathalie Cammas (France Telecom. R&D), Isabelle Amonou (France Telecom. R&D), Nam Hyeong Kim (Korea Univ.), Hyeong Min Nam (Korea Univ.), Haechul Chou (ETRI), Marta Karczewicz (Nokia), Lu Yu (Zhejiang Univ.), Detlev Marpe (Fraunhofer HHI), Arturo A. Rodriguez (Scientific-Atlanta), Wade K. Wan (Broadcom), Stefano Tubaro (Politecnico di Milano), Kue-Hwan Sihn (Samsung Electronics), Hiroshi Watanabe (SC29 / Waseda Univ.), Mohammed Z. Visharam (Sony Electronics), Yotam Toib (Qnix Tech.), Ran Bonlh (Qnix Tech.), Yoshiyuki Yashima (NTT), Woo-Cheol Sung (Pixtree), Yi-Jen Chiu (Intel), Mukta Kar (CableLabs), Caspar Horne (Metta Tech.), Wladyslaw Skarbeck (Warsaw Univ. Tech), Tomokazu Murakami (Hitachi), Chong Soon Lim (Panasonic), Lianhuan Xiong (Huawei), Chen Ying (Thomson), Cristina Gomila (Thomson), Jill Boyce (Thomson), Woo-Shik Kim (Samsung AIT), Hyun Mun Kim (Samsung AIT), Ajay Luthra (Motorola), Thiow-Keng Tan (NTT DoCoMo), Takahiro Kimoto (NEC), Mark Veltman (Sony), Yuriy Reznik (Qualcomm), Phoom Sagetong (Qualcomm), Jong-Ki Han (Sejong Univ.), Lihua Zhu (Thomson R&D Beijing), Hsiu-Hao Chou (OES/ITRI), Kenji Takita (Panasonic), Takeshi Chujoh (Toshiba), Yoshihisa Yamada (Mitsubishi Electric), Pawel Pietrowski (TPSA/CBR), Walt Husak

132

(Dolby Labs), Ilhong Shin (KAIST), Takuyo Kogure (Univ. Tokyo), Michael Isnardi (Sarnoff), Greg Cook (Thomson), Haoping Yu (Thomson), José Roberto Alvarez (Mobilygen), Pankaj Topiwala (FastVDO), Tukihiro Bandoh (NTT), Yolanda Prieto (Freescale), Tomoyaki Yamamoto (Sharp), Sam Narasimhan (Motorola), K. Kalyan Kunal [approx. – illegible] (Samsung), Hak Kwang Kim (Sejong Univ.), Jie Jia (Sejong Univ.), Joern Ostermann (Univ. Hannover), Jan Bormans (IMEC), Luis A. S. Cruz (DEEC – PT), Aljoscha Smolic (Fraunhofer HHI), and Karsten Müller (Fraunhofer HHI).

4 MPEG-7

4.1 MPEG-7 related work in Poznan The MPEG-7 breakout group was active during the whole week. Input documents as reviewed are listed in the table below.

12164 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 15938-7:2003/FDAM 1 [SC 29 N 6756]

12165 SC 37 via SC 29 Secretariat SC 37 NP on Biometric data interchange formats -- Face identity data

12189 SC 29 Secretariat Summary of Voting on ISO/IEC TR 15938-8:2002/DCOR 1

12202

Weon-Geun OhIk-Hwan ChoA-Young ChoHyun-Mi KimDong-Seok JeongHae-Kwang KimSung-Phil Heo

Feasibility Test of MPEG-7 Visual Descriptors as a visual identifier descriptor

12207 Wo Chang Propose Revised of WD 1.0 Photo Player MAF

12211 Soo-Jun Park Report of cross verification result of Image Categorisation into classes-VCE2

12212 Soo-Jun Park Report of cross verfication result of Identity-based clustering, indexing and retrieval of images- VCE-3

12235Sang-Ki KimSangyoun LeeSun-Jeong Kim

Propose Thermal Image Sensor for VCE-4

12255Sangyoun LeeSun-Jeong KimJung-Sup Ryu

Report of cross verification result - VCE-2

12256Sangyoun LeeJung-Sup RyuKim

Report of cross verification result- VCE-3

12274

Sang-Kyun KimSeungji YangJun Ho ChoYong Man RoJi-Yeon Kim

A MAF Photo Player for Enhanced Photo Browsing

12275

Sang-Kyun KimSeungji YangKyong Sok SeoYong Man RoJi-Yeon Kim

CE Report for Photo Categorization (VCE-2)

12276

Sang-Kyun KimKyong Sok SeoSeungji YangYong Man RoJi-Yeon Kim

Test Dataset for Person-identity-based Photo Clustering Experiment (VCE-3)

12277 Sang-Kyun Kim CE Report for Person-identity-based Photo Clustering

133

Seungji YangKyong Sok SeoYong Man RoJi-Yeon Kim

Experiment (VCE-3)

12283 Akio YamadaSang-Kyun Kim MPEG-7 Visual part of eXperimentation Model Version 24.1

12284 Akio YamadaSang-Kyun Kim WD 2.1 of ISO/IEC TR 15938-8:2003/Amd.2

12285 Akio Yamada A Preliminary Study on the addition of a file format for Image Player

12290 Akio YamadaSang-Kyun Kim Editors' Study of ISO/IEC 15938-3:2002/FPDAM 2

12296 Akio YamadaR. O'Callaghan Editors's Study of ISO/IEC 15938-8:2002/DCOR 1

12312 Karol WnukowiczWladyslaw Skarbek

Dominant Color Temperatures - a new descriptor or an application of the Dominant Color descriptor

12330 Leszek CieplinskiRobert O'Callaghan

UK NB comments on the text of ISO/IEC 15938-3:2002/FPDAM 2

12331 Robert O'CallaghanMiroslaw Bober A Face Recognition Dataset from Video Content

12353Herbert Thoma Stefan Krägeloh Harald Fuchs

Application Scenarios and Derived Requirements for Video Player MAF

Work continued on Visual tools and Description Schemes for image or photo- libraries.The main objective is to develop new visual Description Schemes, Descriptors and other algorithms for applications related to digital image libraries, such as personal collections of photos from digital cameras. Tools supporting the situation-based image clustering has been completed and now work focuses on Image Categorisation, Identity-based image clustering, and Robust Face Recognition technology.

A new non-normative tool called Dominant Colour Temperature was approved for XM (M12312). The Dominant Color temperature is an extension of Dominant Color capable of characterizing images with multiple regions. It is a non-normative tool which extracts meta-data from the dominant colour descriptor. Dominant Color Temperature supports query by example or query by value and image browsing based or multiple regions and their color temperatures.

A significant progress was made in further development of the Photo Player MAF. A joint meeting on Photo Player MAF was held with Requirements and the PP application requirements were reviewed and further refined. Working Draft (v. 2.0) of the Image Player MAF was completed. Request for subdivision of ISO/IEC 23000-3 MPEG Photo Player Application Format was issued.

Technology introduction documents were created for Low-level descriptors, High-level descriptors and Photo Player MAF.

Study Text of Part-3 FPDAM2 was issued.Text of Part-8 PDAM2 was issued.

New Core Experiments were defined/approved:VCE-2 - Image CategorisationVCE-3 – Identity-based clustering, indexing and retrieval of imagesVCE-4 – Robust Face Recognition technology

134

VCE-5 – Evaluation of MPEG-7 Face Recognition Technology on IR Images

4.2 Output documents related to MPEG-7 Visual

No. Title TBP Available15938-3 MPEG-7 Visual

7316 Study Text of ISO/IEC 15938-3:2002/FPDAM2 No 05/07/297317 MPEG-7 Visual XM Document version 25.0 No 05/07/297318 Description of Core Experiments for MPEG-7 New Visual Extensions No 05/07/29

4.3 Output documents related to MPEG-7 Part 8

No. Title TBP Available15938-8 MPEG-7 TR (ext. and use of MPEG-7 descriptions)

7321 Text of ISO/IEC TR 15938-8:2002 COR1 No 05/07/297322 Text of ISO/IEC TR 15938-8:2002 PDAM2 No 05/07/29

4.4 Output documents related to Photo Player MAF

No. Title TBP Available23000-3 Photo Player Multimedia Application Format

7323 Request for Subdivision of ISO/IEC 23000: Part 3 Photo Player Multimedia Application Format

No 05/07/29

7324 WD 2.0 of 23000-3 Photo Player Multimedia Application Format No 05/07/29

5 New Work Item: 23002 Video Technologies

Request for a new work item "ISO/IEC 23002 MPEG Video Technologies" was made in N7516. The first purpose of this new work item is specification of MPEG video coding technologies that could have references with different MPEG standards. As a first component, it is planned to include the specification for IDCT accuracy requirements, which would need to be referenced by 11172-2, 11172-4, 13818-2, 13818-4, 14496-2 and 14496-4. Future projects could be a common definition of color space code points for different standards, and usage of video coding tools across different standards. To simplify the specification of relationships between or across MPEG standards, such work is better performed within a framework that is independent of the already existing standards.

No. Title TBP Available23002 MPEG Video Technologies

7516 Request for New Work Item ISO/IEC 23002 MPEG Video Technologies No 05/07/29

6 Explorations

6.1 Multiview Video Coding

In the preparation of the Poznań meeting, the AHG performed AVC anchor encoding of the different test sequences (with camera parameters provided) that had been made available by the time of the Busan meeting. Extensive viewing of the decoded results was performed for final selection of the test cases in the CfP. In total, 8 sequence sets with a variety of camera configurations were selected. It was furthermore found, that camera parameters for some of the

135

sequences did not match, showing up to 19 pixel deviation from the expected correspondences. During the week, further investigations were made to clarify this issue; camera parameters for one sequence were already corrected with better precision, and providers of other test sequences were contacted which gave confirmation that most probably all necessary information is available to provide parameters with sufficient precision within three weeks after the meeting.As a consequence, it was decided that the CfP on Multiview Video Coding could be issued (N7327). Besides the precise description of the test cases and rate points, many improvements of objectives, evaluation section and logistics section were made. Due to the fact that camera parameters are expected to be urgently needed by some of the expected proposals, the original timeline of the draft call (submission by September) appeared to be too tight. It was therefore decided to issue the CfP with the following timeline: Pre-registration October 10 Final registration November 28 Submission of decoded sequences December 12 Tests performed beginning of January 2006 Evaluation during January meeting

Documents reviewed:a) Related to the CfP and test sequences

12198

Yoshinori SugiharaTsuyoshi MoriyamaHideaki KimataMasaki Kitahara

Keio test data sets for MVCCould be useful for later investigations, but too late for tha CfP, because no anchor encodings available. Thank providers in resolutions.

12214 Masaki KitaharaHideaki Kimata

Results on AVC Reference Coding of KDDI Test Sequences for Call for Proposals on Multi-View Video Coding

12215 Masaki KitaharaHideaki Kimata

Results on AVC Reference Coding of Tanimoto Laboratory Test Sequences for Call for Proposals on Multi-View Video Coding

12230

Alexander Behrens Emin Martinian Huifang Sun Anthony Vetro Jun Xin

Comments on Call for Proposals on Multiview Video CodingProblem in definition of camera position, translation vector (relating to world coordinate system) should be interpreted consistently. Deviation in HHI 19 pixels vertically, in Microsoft 9 pixels vertically.

12231

Alexander Behrens Emin Martinian Huifang Sun Anthony Vetro Jun Xin

Multiview Video Anchor Coding Results for Ballroom and Exit

12239 Hideaki KimataMasaki Kitahara

Usage and description of camera parameters for Multi-view Video Coding and Free-viewpoint TVCamera parameters as presently used (SMPTE) may not be the best representation (global world coordinates, but local world coordinates would be more appropriate). SMPTE parameters are in principle precise enough, but may be mis-interpreted. Other representations, such as pixel-based, may be more appropriate. For applications, users should be provided with a good interface (e.g. layout scheme of available camera positions) that makes view adaptation easy.

12240Yung-Lyul LeeYung-Ki LeeDae-Hee Kim

Anchor coding results of HHI sequences for Call for Proposal on Multi-view video coding

12252 Fons Bruls Results on AVC Reference Coding of Tanimoto Laboratory Test Sequences for 3DAV CfP

12280 Daehee Kim Anchor bitstreams for Call for Proposals on multi-view video

136

Namho HurSoo In Lee coding

12298Aljoscha Smolic Philipp Merkle Karsten Mueller

Anchor Coding Results for Fraunhofer HHI Test Sequences

12328

Kwanghoon SohnYongtae KimChangseob ParkJaeho Lee

Comments on requirement and camera parametersModify requirement on view generation "support" ->"enable", add "efficient". Proposal on camera parameters is technical solution that anticipates a possible response to the Call.

12338

Masayuki TanimotoToshiaki FujiiTakanori SenohTerumasa AokiYoshinori Sugihara

Test Sequences with Different Camera Arrangements for Call for Proposals on Multiview Video Coding

12365

Cristina GomilaYeping SuAntonio OrtegaJae Hoon Kim

Comments on the Preliminary CfP on Multi-view Video CodingEncourage use of different camera settings, different spatial and temporal sampling, but should be more aligned with application perspectives (registered more important). Bit rate constraint should be defined on all views. Change of non-normative elements in H.264 should be kept minimal to minimize the influence on the results. In general, complexity analysis is very important. Add sentence on scope of the Call, but not concrete requirement on bitrate saving. Do not request proponents to submit single-view results." Should provide flexible quality adaptation for different views". Add more specific on complexits (amount of memory, memory accesses, number of ops per reconstructed pixel etc.)

12402 Akio Ishikawa KDDI Test Sequences for Call for Proposals on Multi-View Video Coding

b) other technical input

12270

Hyok SongByeongho ChoiJinwoo BaeJisang Yoo

A framework for Multi-view Video Coding using Accumulated image

12278

Yo-Sung HoSeung-Uk YoonEun-Kyung LeeSung-Yeol KimSeung-Hwan KimKugjin YunDaehee Kim

Intermediate Result on Multi-view Video Coding using Layered Depth Images

12301

Philipp Merkle Aljoscha Smolic Karsten Mueller Thomas Wiegand

Statistical Evaluation of Spatiotemporal Prediction for MVC

Output documents:

No. Title TBP Available3D AV

7327 Call for Proposals on Multi-View Video Coding Yes 05/07/29

6.2 Video Coding Tools Repository

137

The VCTR breakout group was busy throughout the week. As a next major step, a demo of VCTR (including software implementation) is planned for the October meeting. The following demonstrations were discussed and will further be prepared within the work of the AhG:

1. Reconfigurable Coding: Encoding is carried out freely or optimally employing either one of three combinations of FUs such as MPEG-2, MPEG-4 and AVC. Switching is done at 1: frame level, 2: macroblock level, or 3: bitstream level. The expected benefits are: Efficient reuse of source codes – to show how much % of source codes are shared within reconfigurable decoder, which implies how efficient we can build new decoders. Reference Models need not to take care about MPEG phases, even non-MPEG tools could be studied using VCTR. Eventually, reconfigurable coding may work better than any other single-mode coding, by choosing optimal coding tools from different standards dynamically.a) 1-1 Frame level Reconfigurable coding: Bitstreams for reconfigurable decoder are

generated off-line (or on the fly). Reconfigurable decoder can be built incrementally on the fly. Decoding (playback) can be shown incrementally as well

b) 1-2 Macroblock level Reconfigurable coding: All based on intra-only MPEG-4 SP configuration. Signal this configuration by setting video_object_layer_verid to 6~9 (reserved value). Remove or add some symbols in the SynP FU. Option 1: Replace MPEG-4 DCR FU with MPEG-2 DCR FU. Option 2: Option 1 + MPEG-4 DC/AC prediction. Option 3: Replace MPEG-4 VR FU with AVC SP+VR FU. Option 4: Extend SP+VR FU to support 8x8 luma prediction modes and 8x8 block-based processing. Option 5: Option 4 + new prediction.

c) 1-3 Bitstream level Reconfigurable coding: One bitstream ={ AVC, MPEG-2, AVC}. Flexible configuration, different codecs, different resolutions, different frame/bit rates

2. Profile-free coding: This is the "Do-whatever-you-want Codec". It is planned to realize an "MPEG-4 My Profile" with flexible selection of tools and flexible syntax support (syntax rearrangement and selection). The expected advantage would be that one can evaluate a codec to find a best mode coding for a specific environment.

3. AVC High 4:4:4 profile vs. “Advanced 4:4:4 profile” in VCTR framework: This will be shown for an example of an AVC High 4:4:4 profile decoder (operated intra-only). VCTR Textual description and software will be provided for this. It will be extended as an "Advanced" AVC 4:4:4 intra-only profile decoder similar to the methods presently discussed in JVT, made from the above on the fly by incorporating new tools. This will be a live exercise of creating a new profile using VCTR

4. SW/HW co-design based on VCTR framework: An example for this is an IDCT which can be accelerated by Reference Hardware

The new version of the Study of Video Coding Tools Repository V5.0 (N7329) contains the updated software workplan (presently available: MPEG-4 SP, MPEG-4 AVC BP FUs; planned to be available for the demo: MPEG-2 MP, MPEG-4 AVC High Profile), as well as the demo plan. The VCTR Textual Description V4.0 (N7330) includes AVC BP-based FUs (from M12243). The new version VCTR Software V3.0 is available as N7331.

Documents reviewed:

12242

Kazuo SugimotoYoshihisa YamadaKohtaro AsaiTokumichi Murakami

Contribution to VCTR Reference Software - Intra Only Decoder for AVC Baseline Profile

12243 Kazuo SugimotoYoshihisa YamadaKohtaro Asai

Proposed update of TD 3.0

138

Tokumichi Murakami

12399

Sunyoung LeeHyunsoo AhnHyungyu KimChungku YieEuee S. Jang

Reconfigurable Syntax Description for VCTR

12400

Hyungyu KimSunyoung LeeHyunsoo AhnChungku YieEuee S. Jang

Analysis of Python Language for VCTR Implementation

12401

Hyunsoo AhnSunyoung LeeHyungyu KimChungku YieEuee S. Jang

Status Report of VCTR Reference Software

Output Documents:

No. Title TBP AvailableVideo Coding Tools Repository

7329 Study of Video Coding Tools Repository V5.0 No 05/07/297330 VCTR Textual Description V4.0 No 05/07/297331 VCTR Software V3.0 No 05/07/29

6.3 Wavelet Video Coding

Input documents related to previous exploration experiments were reviewed, and showed some progress in compression quality and/or quality improvement by using different subsampling filters and postprocessing. Starting from this, the next round of exploration experiments was defined in N7333, targeted to show progress in wavelet video area by

• Improved coding efficiency and extended range of SNR scalability • Improvements of intra coding, bidirectional motion estimation, de-blocking/de-ringing

For the Nice meeting, it is planned to conduct a visual performance comparison of t+2D architecture and 2D+t+2D architecture (non multilayer), and also compare against the Palma results (from October 2004). The breakout group also discussed advantages of wavelets under specific application constraints, in particular in application cases with the following requirements:

• no predefined scalability range • high number of spatio-temporal decomposition levels• Applications needing extremely fine grain SNR scalability • For the future, an important target will be to better adapt the evaluation of wavelet

technology for more realistic application scenarios (including appropriate test material).One more output document contains the description of the algorithm and the software which are presently used in the wavelet exploration activity (N7334).

Documents reviewed:

12176Vincent BottreauGrégoire PauJizheng Xu

Vidwav evaluation software manual

12286Ruiqin XiongJizheng XuFeng Wu

Responses to Vidwav EE1

139

12303 Grégoire Pau Maria TrocanBéatrice Pesquet-Popescu Bidirectional Joint Motion Estimation for Vidwav Software

12339

Xiong Xiangyang Ji Dongdong Zhang Jizheng Xu (Microsoft Research Asia) Grégoire Pau Maria Trocan (ENST) Vincent Bottreau (INRIA)

Vidwav Wavelet Video Coding Specifications

12374 Markus Beermann Joint reduction of ringing and blocking for VidWav

12376 Yongjun WuJohn Woods Aliasing reduction for subband/wavelet scalable video coding

12410 Soroush GhanbariLeszek Cieplinski Results of Vidwav Exploration Experiment 3

Output Documents:

No. Title TBP AvailableWavelet Video Exploration

7333 Description of Exploration Experiments in Wavelet Video Coding No 05/07/297334 Wavelet Codec Reference Document and Software Manual V1.0 No 05/08/19

6.4 New IDCT Specification Work

The prior call for proposals on fixed-point 8x8 IDCT and DCT was reviewed in detail, modified as necessary, and finalized for approval. One issue remained open at the time of the meeting, which was the rounding method to be applied in computation of the idea IDCT results. It was agreed that this issue would be finalized by email in the relevant AHG work following the meeting by 15 September 2005.

Documents reviewed (with summary of contents and discussion):

12322 Qi Wen Gao Debin Zhao Siwei Ma Liang Wang Long Xu

High Accurate and multiplierless fixed-point DCT for CFP on 8x8 fixed-point IDCT and DCT standardsApproach based on "AAN" method - (5 multiplies, 8 scaling multiplies) or 47 adds and no multiples plus scaling.Does not include IQ -- separated design as envisioned in IEEE 1180.

12324 Qi Wen Gao Debin Zhao Siwei Ma Liang Wang Long Xu

An example of fixed-point IDCT for CFP on fixed-point 8x8 IDCT and DCT standardApproach based on "Chen" method. Different consideration of scaling. 76 adds and no multiplies (remark: maybe 88).Advantage relative to M12322 is to leave scaling part untouched.

12359 Ci-Xun ZhangLu Yu

Fixed-Point 8x8 IDCTConsiderations: Mostly agrees with CfPAdvocates only including the IDCT (not IQ) in the scope of the CfP effort and subsequent new standard. This approach is not fully agreed by the group, but might be a good solution if determined so with further analysis. (Best to be open to additional information and to be willing to consider other approaches as well before making a decision.)Particular approach outlined. Approach based on "LLM" method (12 multiply, 32 add) or matrix multiply and also with multiply-free.

Output Document:

140

No. Title TBP AvailableFixed-point Implementation of DCT/IDCT

7335 Call for Proposals on Fixed-point 8x8 IDCT and DCT Yes 05/07/29

6.5 Workshops on Future Directions of Video Compression

Based on the trends identified in Busan and evaluation thereafter, a second full day workshop will be held on Sunday, 16 October, 2005 in Nice, France. A Call for participation was issued (N7337). It is expected that the programme will be available by beginning of September, such that it could be included in the final version of the Call.

Output documents:No. Title TBP Available

Workshop on Future Directions in Video Compression7337 Second Workshop on Future Directions in Video Compression – Call

for ParticipationYes 05/09/01

7 Promotion

7.1 Introduction Documents

The Video subgroup approved introduction documents related to the different standards under their responsibility, as well as most relevant exploration work.

Output documents:

No. Title TBP AvailableIntroduction Documents

7290 Introduction to MPEG-1 Video Yes 05/07/297294 Introduction to MPEG-2 Video Yes 05/07/297297 Introduction to MPEG-4 Part 2 Video (rectangular) Yes 05/07/297298 Introduction to MPEG-4 Part 2 Video (arbitrary shape) Yes 05/07/297314 Introduction to MPEG-4 Part 10 Advanced Video Coding Yes 05/07/297315 Introduction to SVC Extension of Advanced Video Coding Yes 05/07/297319 Introduction to Low-level Visual Description Tools Yes 05/08/197320 Introduction to High-level Visual Description Tools Yes 05/08/197325 Introduction to Photo Player Multimedia Application Format Yes 05/08/197326 Introduction to Accuracy Requirements for 8x8 IDCT Yes 05/08/197328 Introduction to Multi-View Video Coding Yes 05/07/297332 Introduction to Video Coding Tools Repository Yes 05/07/297336 Introduction to Fixed-point Implementation of DCT/IDCT Yes 05/08/19

Annex 9Audio

141

Source: S. Quackenbush, Chair, Audio Subgroup

1 Opening of the meeting.........................................................................................................1442 Administrative matters...........................................................................................................144

2.1 Approval of previous meeting report 1442.2 Approval of agenda and allocation of contributions 1442.3 Communications from the Chair 1442.4 Joint meetings 1442.5 Received National Body Comments and Liaison matters 1442.6 Task Groups 144

3 AhG meetings........................................................................................................................1443.1 Spatial Audio Coding (Sunday 0900 – 1300)1453.2 Exploration on Scalable Audio and Speech Coding (Sunday 1400 – 1700) 1463.3 Symbolic Music Representation (Sunday 1400 – 1900) 147

4 Audio plenary, joint meeting and task group activities.........................................................1474.1 Review of AHG reports 1474.2 Received national body comments and liaison matters 1484.3 Audio opening plenary discussions 1484.4 Joint Meetings 149

4.4.1 Joint with MDS, Req on Protected Music MAF......................................................................................1494.4.2 Joint with Req proposed profiles.............................................................................................................151

4.5 Task Group discussions 1514.5.1 MPEG-7...................................................................................................................................................1514.5.2 Spatial Audio Coding..............................................................................................................................1524.5.3 Music Player MAF..................................................................................................................................1534.5.4 Lossless Coding.......................................................................................................................................1534.5.5 MPEG-4 Audio........................................................................................................................................1544.5.6 MPEG-4 Conformance............................................................................................................................1554.5.7 Symbolic Music Representation – Giorgio Zoia.....................................................................................155

4.6 Audio closing plenary discussions 1555 Meeting deliverables..............................................................................................................155

5.1 Press statement 1555.2 Dispositions of Comments 1565.3 Responses to Liaison and NB comments 1565.4 Recommendations for final plenary 1565.5 Establishment of Ad-hoc Groups 1565.6 Approval of output documents 156

6 Future activities.....................................................................................................................1566.1 Schedule of future meetings 1566.2 Agenda for next meeting 1566.3 All other business 1566.4 Closing of the meeting 156

Annex A Participants...............................................................................................................157Annex B Audio Contributions and Schedule...........................................................................159Annex C Task Groups.............................................................................................................164Annex D Output Documents....................................................................................................165Annex E Agenda for the 74th MPEG Audio Meeting..............................................................167

142

1 Opening of the meetingThe MPEG Audio Subgroup meeting was held during the 73rd meeting of WG11, July 25-29, 2005 in Poznan, PL. The list of participants is given in Annex A.

2 Administrative matters2.1 Approval of previous meeting reportThe 72nd Audio Subgroup meeting report had been previously distributed by e-mail and was approved.

2.2 Approval of agenda and allocation of contributionsThe agenda and schedule for the meeting was discussed, edited and approved. It shows the documents contributed to this meeting and presented to the Audio Subgroup, either in the task groups or in Audio plenary. The Chair brought relevant documents from Requirements, Systems and MDS to the attention of the group. It was revised in the course of the week to reflect the progress of the meeting, and the final version is shown in Annex B.

2.3 Communications from the ChairThe Chair summarised the issues raised at the Sunday evening Chair’s meeting, proposed task groups for the week, and proposed agenda items for discussion in Audio plenary.

2.4 Joint meetingsThe joint meetings with Audio over the course of the week are listed here and are reported on below.

Groups What Where Day TimeReq, Audio, MDS, Systems Protected Music MAF (12197, 12289, 12310, 12335,

12360)Req Tue 1600-1800

Req, Audio MPEG-4 Profile (12372) Audio Wed 1700-1730Req, Audio, MDS, Systems Protected Music MAF Req Thu 1500-1600

2.5 Received National Body Comments and Liaison mattersThe NB Comments and Liaison documents for the meeting that require a response are as shown below.No. Title Response by12263 USNB Contribution: MPEG-4 ALS issues S. Quackenbush12264 USNB Contribution: Complexity issues regarding Audio tools S. Quackenbush

12269 SGNB comment on request to extend the work schedule for Spatial Audio Coding Chong Kok Seng

12271 Korean NB's Comment on Spatial Audio Coding Korean HOD12394 Korean NB Contribution: Public MPEG Software issues S-W Kim12348 FNB Contribution: Spatial Audio Coding Activity issues Pierrick Philippe

12150 Liaison Statement from ITU-R SG 6/WP 6A S. Quackenbush12151 Liaison Statement from ITU-R SG 6/WP 6A S. Quackenbush

2.6 Task GroupsTask groups were convened for the duration of the MPEG meeting, as shown in Annex C. Results of task group activities are reported below.

3 AhG meetingsIn order to obtain a compete view of the business of this MPEG meeting, discussion of contributions in AhG meetings are recorded in this section of the Audio report.

143

3.1 Spatial Audio Coding (Sunday 0900 – 1300)Werner Oomen, Philips, presented12226 Werner Oomen

Francois MyburgAdditional subjective test results on High quality operation modes of spatial audio coding RM0.

This showed the results of subjective tests which indicate that high side information rates of 32 kb/s to 192 kb/s lead to subjective performance that approaches transparency. Specifically, a parameterization of AAC-LC core at the rate of 128 kb/s plus 192 kb/s spatial side information produces a subjective score that is statistically indistinguishable from AAC-LC discrete multi-channel coding at 320 kb/s. The additional side information rate is largely used to code the “residual signal” representation. The Chair recommended that these new subjective test results be added to the previously RM0 verification results to make a new output document that could used to document all performance data on spatial audio coding.Juergen Herre, FhG, presented

12368

A. HoelzerW. OomenK. KjörlingJ. Herre

Additional subjective test results on R/D scalability of Spatial Audio Coding RM0

This contribution explored scaling spatial side information below those rates previously documented in the RM0 verification tests, specifically, from 6.6 kb/s down to 1.8 kb/s. Jean-Bernard Rault noted that the lowest rate, 1.8 kb/s delivered distinctly worst performance for the fountain music item. Juergen Herre indicated that this was due to the fact that this test item was significantly different from the others in that it is really composed of three unrelated signals that are mapped to C (fountain), L,R (piano) and Ls,Rs (birds). As with the previous contribution, the Chair recommended that these results be captured in an output document.Chong Kok Seng , Panasonic Singapore Labs, presented

12267

Chong Kok Seng Takeshi Norimatsu Yoshiaki Takagi Neo Sua Hong

Request for formal verification test of the Low Complexity mode of Spatial Audio Coding RM0 and backward compatibility with LP-SBR

This contribution makes three requests: Verify performance of the Low Complexity mode of Spatial Audio Coding Explore how to modify Spatial Audio Coding so that it can be tightly integrated into the

Low Power mode of the HE-AAC coder. Explore whether a low complexity version of Spatial Audio Coding is possible (i.e. using a

real filterbank)

The Chair noted that the first request can be added to a workplan for Spatial Audio Coding, and that the second two requests really requires a core experiment (for real processing) within the Spatial Audio Coding work.Kristofer Kjörling, Coding Technologies, presented

12320

Kristofer KjörlingWerner OomenFrancois MyburgJürgen Herre

Proposed corrections and clarifications to the WD on Spatial Audio Coding

This contribution presented corrections to errors and ambiguous text in the Spatial Audio Coding WD text. All changes are editorial. It was the consensus of the AhG to take the contribution document as the new WD text.Kristofer Kjörling, Coding Technologies, presented12361 Kristofer Kjörling Proposal for extended signalling in Spatial Audio

This proposes a method of compactly coding the geometry of a tree comprised of a series of “One-to-Two” (OTT) splits such that Spatial Audio Coding can support a mapping of an arbitrary number of transmitted channels and output channels. It was the consensus of the AhG to proceed with this work as a core experiment, but without the need for rigorous listening tests or the need to show “significant compression improvement”. David Virette, France Telecom, presented

144

12350David ViretteManuel BriandPierrick Philippe

Comments on the Spatial Audio Coding RM0 and Working Draft

This contribution presented a number of editorial corrections to the WD. It also noted that a number of functionalities of RM0 were not tested in the RM0 verification test. It was the consensus of the AhG that the testing of the following functionalities should be addressed in a workplan

Artistic downmix Matrix Surround compatibility Temporal Envelope Shaping (TES)

The Chair noted that the last, TES, will be addressed when the performance of LC SAC is tested. In Yong Choi, Seoul National University, presented

12265

In Yong ChoiSang Bae ChonJeongil SeoInseon JangKoeng-Mo SungKyeongok Kang

Objective Measurement of Total Auditory Quality of Spatial Audio Coding

This contribution presented a method for characterizing the performance of Spatial Audio Coding algorithms using objective measures, specifically ITU-T PEAQ. The Chair encouraged Seoul National University to continue to work with MPEG to develop this new “multi-channel” version of PEAQ.Werner Oomen, Philips, presented12227 Jeroen Breebaart

Christof Faller Core experiment on non-guided spatial audio coding

This contribution proposes a core experiment for spatial audio coding based on zero spatial side information. In this case the spatial parameters are derived directly from the time/frequency characteristics of the transmitted channels. Optionally, some time/frequency tiles can be “non-guided” (zero side information), while others have spatial side information as in the current WD. Subjective test results in the contribution indicate that the performance of non-guided Spatial Audio Coding is better than Dolby Prologic II and Logic 7 at the 95% level of significance.The AhG recommends proceeding with this core experiment.

3.2 Exploration on Scalable Audio and Speech Coding (Sunday 1400 – 1700)Andy Tescher, Microsoft, presented12262 Jin Li

James D. JohnstonA Progressive to Lossless Embedded Audio Coder (PLEAC)

The salient features of the PLEAC technology are: Scalable bitstream, to the granularity of one byte The reconstruction fidelity is commensurate with bitrate, up to the capability of lossless

reconstruction. Both the bitstream scaling engine and the decoder have low complexity.

The presenter emphasised that Microsoft is happy to contribute the source code for this technology to a potential MPEG work item, and would be happy to collaborate with other partners on advancing the technology. Additionally, he felt that the most appropriate next steps for the CfI would be continued exploration, e.g. using this technology. That may include performance evaluation in the context of a specific use case scenario.Kevin Short, Chaoticom Technologies, presented12253 Kevin M. Short KOZ scalable audio compression

He gave a brief overview of the products and markets of Chaoticom, and then gave an overview of the KOZ technology. One of the virtues of the technology is its additive synthesis representation, which permits scaling of the bitstream rate with commensurate scaling of signal quality. It is also a low complexity decoder (approximately 60 MIPS). The encoder runs at or slightly higher than real-time. Anisse Taleb, Ericsson, presented

145

12311 Anisse TalebScalable Speech and Audio Coding based on the 3GPP AMR-WB+ standard: Response to the CfI on scalable speech and audio coding

This contribution addresses the use case of audio distribution over the mobile phone wireless channel. The signal of interest is any of speech, music or speech over music. The important requirements for a successful codec that addresses this market are:

Unification of speech and audio coding Capability of low bitrate representation (i.e. high compression)

Enhanced AMR-WB+ appears to satisfy these requirements in that it Offers multiple coding models Can select between the models base on an objectively computable criterion Supports fine-grain scalability Has low complexity Achieves fine-grain scalability via an extension layer structure in the bitstream.

The Chair noted that the contribution has subjective performance data for AMR-WB+, but not Enhanced AMR-WB+ (so that the performance data is that of Enhanced AMR-WB+ with only the base layer). Eunmi Oh, Samsung, presented

12302 Eunmi OhYang-Seock Seo

Response to CfI on scalable speech and audio coding: The need for a new coding paradigm and its application scenarios

This contribution presented a summary of the market trends in the mobile phone marketplace, with the intent of providing a firmer basis for the mobile network use case. The most important issues are:

Unification of speech and audio coding High compression Low complexity

The presentation concludes that scalability is not a strong need for this market. Scalability certainly can be provided, but without any compromise in coding efficiency and complexity. As a conclusion, a table of compression, complexity and functionality is provided.

DiscussionKevin Short noted that many mobile phone operators are looking for new codecs for their recently deployed 3G systems. Anisse Taleb commented that as 3G services become successful, those links will become congested, and scalability may become more important.

3.3 Symbolic Music Representation (Sunday 1400 – 1900)The following documents were reviewed. See the SMR task group section of this report.

12223Giorgio ZoiaPaolo NesiPierfrancesco Bellini

Draft Core Experiment methodology for MPEG-4 Symbolic Music Representation

12236

Hyoung-Joong KimYong-Soo ChoiYong-Ju ChoJae-Gon Kim

The Extended Symbolic Music Representation for Korean Traditional Music.

12279Pierfrancesco BelliniPaolo NesiMaurizio Campanai

MPEG-4 SMR RM0 documentation for SMR WD text

4 Audio plenary, joint meeting and task group activities4.1 Review of AHG reportsThere were no requests to review any of the AHG reports.

146

4.2 Received national body comments and liaison mattersThe NB comments and liaison documents listed in Section 2.5 were presented and discussed. The Audio Subgroup had consensus on the following

USNB Contribution: MPEG-4 ALS issues – an item will be included in the workplan on ALS to check cross-platform performance. Specifically, Audio Research Labs will decode ASL bitstreams that use the RLS-ALS predictor on an Apple G4 platform and confirm lossless reconstruction.

USNB Contribution: Complexity issues regarding Audio tools – the Audio Subgroup will take this under consideration for new work, i.e. for work that is at or earlier than WD stage.

SGNB comment on request to extend the work schedule for Spatial Audio Coding – this needs additional discussion.

Korean NB's Comment on Spatial Audio Coding – this needs additional discussion. Korean NB Contribution: Public MPEG Software issues – it appears that much reference

software is on the SC29 web site, but not referenced on the ISO URL indicated in the NB comment (Chairs note – after discussion later in the week it appears that this is not the case). Furthermore, some reference software is only in the MPEG-4 VM, and must be added to the 14496-5 as an amendment. A workplan will be drafted to address these issues.

FNB Contribution: Spatial Audio Coding Activity issuesSchedule of Spatial work item – this needs additional discussion.Testing the performance of the following will be done as indicated here:

o TES – Temporal Envelope Shaping – to be done as part of a verification of Spatial LC mode.

o Residual Coding – has already been done in the 32 kb/s side information results.o Arbitrary DownMix – will be done as soon as possible, but balanced with the

priority of the other tasks in the Spatial work. Liaison Statement from ITU-R SG 6/WP 6A - WG11 can give SG6 a wealth of subjective

assessment data. The Audio Subgroup Chair will correspond directly with the WP 6A Rapporteur to coordinate the transfer of this information.

Liaison Statement from ITU-R SG 6/WP 6A – WG11 can give SG6 a wealth of multi-channel subjective assessment data. The Audio Subgroup Chair will correspond directly with the WP 6A Rapporteur to coordinate the transfer of this information.

4.3 Audio plenary discussionsMondayThe Chair reminded delegates that all AhG consensus positions or explicit recommendations from Sunday’s AhG meetings or from the AhG report are recommendations to the Audio Subgroup. The Chair asked delegates for approval of all such positions and recommendations (e.g. as documented in this report), and it was the consensus of the Audio Subgroup to do so.The Chair presented

S QuackenbushFeature Extraction plugin API: http://feapi.sf.net

This is a model for what MPEG-7 reference code could be. Kristofer Kjörling, Coding Technologies, presented

12319Kristofer KjörlingWerner OomenJürgen Herre

MPEG Surround, a proposed new name for the Spatial Audio Coding tool

The group was generally supportive of this new name. However there was some concern that the name was already used to refer to Layer II BC in the context of DVD. The Chair will consult with the Convenor. After seeking the opinion of the Convenor (which was supportive of the new name), and reporting that back to the Audio Subgroup, it was the consensus of the group to adopt the name MPEG Surround for this work item.The Chair presented

147

12347 Schuyler Quackenbush Proposed Tutorial on MPEG Surround Audio Coding

This is a vehicle for publicizing MPEG technology, and the intension is to turn this into a public WG11 document and post it on the MPEG tutorials web page: http://www.chiariglione.org/mpeg/tutorials.htmPaolo Nesi, XXX, presented

12223Giorgio ZoiaPaolo NesiPierfrancesco Bellini

Draft Core Experiment methodology for MPEG-4 Symbolic Music Representation

This is very similar to the MPEG-4 CE doc. One figure of merit for proposed technology will be similar to the one use in the evaluation process for the SMR CfP, another will be subjective assessment. ather than subjective performance data as was used in MPEG-4.WednesdayThe Chair reported on the Tuesday evening Chairs meeting. The Convenor is quite supportive of the name “MPEG Surround” despite reported issues (e.g. Layer II BC). It is the consensus of the Audio Subgroup to rename the Spatial Audio Coding work item MPEG Surround.The Chair reported that if the MPEG Surround work is promoted to CD at the Wednesday plenary of the 74th meeting, then it could be promoted to FCD at the Friday plenary of the 75th meeting, although ballot comments could arrive as late as Thursday of the 75th MPEG meeting. It was the consensus of the Audio subgroup to keep the MPEG Surround work item at working draft for this meeting, and to promote it to CD at the 75th meeting. If there are significant numbers of substantive Core Experiments proposed at the 75th meeting, the group will discuss as to whether to keep the remainder of the timeline (i.e. FDIS in July 2006) as is, or to extend it. ThursdayJan van der Meer, Philips, gave a presentation on the use of Systems to signal MPEG-4 audio. The Profile and level indication is signalled in the MPEG-4 audio descriptor, and signal values are defined in profile and level table in Systems specification. System Target Decoder paramerters : Size of decoder buffer

o Currently 3584 B, except for MPEG-2/4 AACo Leaves no room for multiplexing

Transfer rate buffer (for transport streams)o Specified size of R (in kb/s) for each AOT

AudioConfig() and use of LATM/LOASo MPEG-2 Systems provides no alignment between PES packets and audio

accessUnitso LATM/LOAS is needed to provide synchwordso LATM/LOAS carries audioDecoderConfig “in band”o Could carry audioDecoderConfig and/or profileAndLevel at a “higher level” within

Systems multiplex.

MPEG-2 Systems carries a table enumerating MPEG-4 Audio profile and level. This means that each new Audio technology must amend MPEG-2 Systems.Recommended ActionThe process of adding new audio technology should include these steps:

Start a Systems AMD in parallel with an Audio AMD. Estimate the B and R values for the technology. Review as to whether LATM/LAOS is a suitable transport for these audio streams.

148

4.4 Joint Meetings 4.4.1 Joint with MDS, Req on Protected Music MAF

TuesdayStefan Kraegeloh, FhG, presented12360 Stefan Kraegeloh

Harald FuchsProposed Protected Music Player MAF

This contribution proposes a protected format for the Music Player MAF. It does not address key management or digital rights management. It does propose to specify a specific means of content encryption (AES128) signalling of the encryption scheme means to enable content random access. Content ID for association with a key Key management system URI

Hendry Munchurl Kim, Information and Communications University, presented

12197 HendryMunchurl Kim

Protecting and Governing Music MAF Player Format based Contents by using MPEG-21 IPMP

This contribution proposes that each aggregator/distributor should have the option of his own protection toolsThis contribution was presented:

12289Zhongyang HuangShengmei ShenTakafumi Ueno

IPMP Protection for MPEG-A Music Player

This gives details of how MPEG-21 IPMP can be used to provide protection in the context of Music Player MAF.This contribution was presented:

12310

Jongheum KimSungphil HeoWeongeun OhHaekwang Kim

IPMP metadata protection for super-distribution MAF

This gives details of how a super-distribution protection scheme might work. It clearly needs to support: Integrity Authentication Non-repudiation

DiscussionIt is noted that there are three cases in the Music Player MAF: Single track in MP4 file Single track in MPEG-21 file Multiple Music Player MAF MP4 files in an MPEG-21 file

A possible unification of these four contributions is:Case 1: Add content protection with AES128 as default Signalling within MP4 file Store license information in separate file

Case 2 and Case 3: Protection description and signalling via MPEG-21 IPMP. AES128 could be the default tools. Signalling within MPEG-21 DID

Store license either in DID or in a separate file License in separate file MPEG-21 file format with DID Supports case 1, 2, and 3

149

Florian Pestoni, Microsoft, was encouraged that there is a unified solution from the four proposals, and that it uses some of MPEG’s latest technology. However he was concerned that AES128 was specified as mandatory in some cases. Fernando commented that the Profiles under Consideration currently specify one mandatory encryption algorithm. Florian maintained that the flexibility provided by the IPMP toolbox affords the greatest interoperability. He continues to be very concerned that AES128 is designated as either mandatory or default. Fernando strongly asserted that the essence of a MAF is to make a choice so that interoperability is increased.Zvi Lifshitz, Obtibase, presented

12335Zvi LifshitzCyril ConcolatoTIRAMISU consortium

Proposal for free-distribution MAFRequirements

This contribution proposes to make as many choices as possible concerning a media MAF so as to provide maximum interoperability. It addresses only stored media representations for streaming over the Internet, and is an update of the “Super-distribution” in the MAF under consideration.SummaryIt was decided that next actions would be to:

1. Draft text that could be inserted in MAFs Under Consideration. 2. Review this text in a later joint meeting

ThursdayStefan Kraegeloh, FhG, presented the breakout group text. Much progress was made. Florian Pestoni summarized the position: lack of consensus on the use cases lead to lack of consensus on the requirements. Nils Rump, Content Guard, suggested the name “Protected Music File Format MAF.”

4.4.2 Joint with Req proposed profiles

Ralph Sperschneider, FhG, presented12372 Ralph Sperschneider

Manfred LutzkyProposed new profile for high-quality communication

This had already been presented to the Audio Subgroup. Requirements endorses this profile, and it was the consensus in the joint meeting to either have this profile as a WD or as a CD, but in any case the document would contain only this profile specification. The Chair will discuss in the chairs meeting whether it is more appropriate to issue this as WD or CD. At the Thursday Chairs meeting, the Convenor was supportive of starting this work at CD, and so it was the consensus of the Audio subgroup to do so.

4.5 Task Group discussions4.5.1 MPEG-7

Markus Beerman, presented12241 Holger Crysandt

Markus Beermann Cross check results of the EnhancedAudioSignature

This contribution presented identification scores for the EnhancedAudioSignature (EAS) versus AudioSignature (AS). The results showed that EnhancedAudioSignature provides overwhelmingly better performance as compared to AudioSignature. Three test were presented

1. Nine song simple test.2. Fingerprint metadata for more than 20000 songs. The EAS technology provided 100%

correct recognition versus 12% for AS for a test with more than 20000 trials.3. Fingerprint metadata for more than 3000 songs. EAS provided similar performance.

In addition the cross-check verified that all descriptions are valid XML. This successfully concludes this CE. The Audio Subgroup agrees to promote this technology into the MPEG-7 FPDAM.Gregoire Carpentier, IRCAM, presented12208 Gregoire Carpentier

Jerome barthelemy Core experiment on weighted scales

150

The presentation reviewed the scale descriptor, and then proposes an extension to that descriptor. The proposed extension would capture the relative importance of the various components of a scale, such that the scale can be identified. Matthias Gruhne, Fraunhofer IDMT, presented12367 Matthias Gruhne CE Crosscheck on Weighted Scale Type

This is a cross-check on the CE in contribution 12208. The Audio Subgroup agreed to incorporate this into the FPDAM text.

4.5.2 Spatial Audio Coding

Jeongil Seo, ETRI, presented

12266

Jeongil SeoSeungkwon BeackIn Young ChoiInseon JangKyeongok KangJinwoo HongMinsoo Hahn

An Efficient Quantization Method for CLD

This contribution presented an alternate quantizer and associated Huffman table for the Channel Level Differences (CLD). There was discussion of evidence of performance. It was felt that evidence presented was possibly subject to large errors in the mean performance. Provided in the CE proposal: Information as to whether the new tool is planned to be an addition to the RM or if it is planned to replace a tool in

the RM. A technical description which is accurate enough that the main function can be duplicated by someone skilled in the

art (this does not mean bit-identical replication of results). A one page “executive summary.” A complexity evaluation giving estimates of ROM, RAM and signal processing functions (e.g. MAC) necessary to

implement the tool.

Still to do in the CE: The WD changes (syntax and description) which are necessary to incorporate the proposed technology. Evidence for the merit of the new tool (for example listening test results).

The Audio Subgroup agreed that this will be provided in collaboration with ETRI and RM0 proponents. Subjective test data should be using MUSHRA listening test. The relevant CE procedure can be found in N7140, Revised core experiment methodology for MPEG-4 audio. The Chair noted that, historically in MPEG-4 development, a compression improvement 5% was considered to indicate significant improvemnt, but that this was always subject to consensus of Audio Subgroup, so that every core experiment proposal needs to be evaluated on a case by case basis. Hee-Suk Pang, LG Electronics, presented

12268

Hee-Suk PangDong Soo KimJaehyun LimHYEN-O OH

Subband domain energy compensation for spatial audio coding

This contribution presented a method for a decoder to compensate for excessive level in spectral regions of the dowmix signal. It was noted that this essentially fixes a problem in the encoder-generated downmix signal, and therefore could be done in the encoder. The Chair noted that such a solution is non-normative and hence out of scope for a core experiment. The proponent will bring to the next MPEG meeting information that compares the alternatives of having the encoder do the compensation versus having the decoder do it.Werner Oomen, Philips, presented

12225Werner OomenKristofer KjörlingJuergen Herre

Request for spatial audio coding profiles

This contribution proposes profiles for Spatial Audio Coding based on identified market needs (e.g. Digital Audio Broadcast). The Audio Subgroup agrees to put support for profiling in the standard, and a description and industry support for profile in an output document.

151

4.5.3 Music Player MAF

Jeongyeon Lim, ICU, presented12396 Jeongyeon Lim

Munchurl KimSynchronization of Multiple JPEG data to MP3 tracks in Music MAF Player Format

This contribution proposes to extend the Music Player MAF to support the functionality of a slide show with background music. The Chair noted that it was not clear whether this is more related to the current Music Player MAF, the proposed Camera MAF, or is a new category of MAF.

4.5.4 Lossless Coding

Tilman Liebchen, TUB, presented12327 Tilman Liebchen Additional Information on the Performance of ALS

This contribution presents new compression performance information. In addition, if presents some first ideas on profiles for ALS. Finally, it notes that the Performance and Complexity of Lossless Coding Architectures document is a vehicle to promote the lossless coding, and the current version of this document should be reviewed and revised so as to promote greater clarity, There was some helpful discussion on the proposed profiles. Ralph Sperschneider pointed out that each profile should be an audioObjectType (at least if one follows the model of most other MPEG-4 profiles), and that there is currently only one audioObjectType in the ALS specification. It was the consensus of the Audio Subgroup that it is premature to put any table proposing profiles in an output document, but rather to continue discussion on profiles in the AhG period and bring contributions on this topic to the next meeting, with the possibility of an output document at that next meeting that addresses ALS profiles. The Chair reiterated the consensus position of the last meeting, that being that there should be one profile of ALS that has low computational complexity.Rongshan Yu, I2R, presented

12380

Rongshan Yu Ralf Geiger Xiao Lin Susanto Rahardja

Update on Performance and Complexity of Lossless Audio Coding Architectures

The updates presented in this contribution consist of SLS has reduced ROM requirements New tables showing performance of

o Compression performance of ALS-RLS predictor.o Compression performance of SLS in core mode as compared to ALS with AAC

simulcast.o Reduced memory size of SLS architecture.

The Audio Subgroup agrees to have 12380 turned into an output document, adding performance information from m11896 and m12327. Finally that there be an AhG mandate to start from a “blank sheet of paper” and create a new performance document that is in the spirit of promoting both technologies. The Chair noted that if the AhG does not make sufficient progress on this mandate we will use N12380 as the basis for the performance document. Ralf Geiger, FhG, presented

12381

Ralf Geiger Markus Schmidt Jürgen Herre Thomas Sporer

Proposed Verification Test Methods for MPEG-4 SLS

This contribution presented a methodology for testing the performance of SLS at intermediate bitrates. The method uses tandem coding and PEAQ objective quality assessment. Several tandeming configurations are considered and tandeming performance with PEAQ measurements are given.Rongshan Yu, I2R, presented

12337Rongshan Yu Xiao Lin Susanto Rahardja

Integration of MPEG-4 SLS with MPEG-4 scalable audio coding tools

152

This contribution shows how to integrate SLS with various scalable MPEG-4 base coders. Since SLS can be carried in a separate elementary stream, the proposal does not involve any syntax changes. The only proposed additions to the FDAM text would be semantic information on the processing steps that would lead to lossless reconstruction. The Audio Subgroup agreed to adopt these changes into the FDAM text, with the understanding that proof that lossless reconstruction is possible for the two base codecs will be presented at the next meeting, along with the associated reference code. With the assumption that interested experts have already reviewed this contribution, it was not presented. The Chair noted that the Subgroup can review essentially the same information when it is time to approve output documents.

12379

Rongshan YuRalf GeigerXiao LinSusanto Rahardja

Proposed Text of ISO/IEC 14496-3:2001/FDAM 5, Scalable Lossless Coding (SLS)

Takehiro Moriya, NTT, presented

12244Yutaka KamamotoNoboru HaradaTakehiro Moriya

Proposed corrections to avoid inconsistency in study on FPDAM4 ALS

Tilman Liebchen, TUB, noted that the proposed changes have already been incorporated into the reference code for some time. The Audio Subgroup agreed to adopt proposed changes into FDAM text.Rongshan Yu, I2R, presented12336 Haibin Huang

Wee Boon ChooProposed Corrections to Study on MPEG-4 ALS FPDA M4

The proposal aligns the text with the reference code such that their behaviour under overflow conditions are the same. The Audio Subgroup agreed to adopt the proposed changes into the FDAM text.Tilman Liebchen, TUB, presented12326 Tilman Liebchen Proposed Corrections and Modifications of Study on

FPDAM4 (ALS)

The Audio Subgroup agreed to adopt the proposed changes into the FDAM text.With the assumption that interested experts have already reviewed this contribution, it was not presented. The Chair noted that the Subgroup can review essentially the same information when it is time to approve output documents.

12325 Tilman LiebchenProposed Text of ISO/IEC 14496-3:2001/FDAM 4, Audio Lossless Coding (ALS), new audio profiles and BSAC extensions

4.5.5 MPEG-4 Audio

Ralph Sperschneider, FhG, presented12372 Ralph Sperschneider

Manfred Lutzky Proposed new profile for high-quality communication

This contribution proposes the “AAC Communication Profile” which contains the ERAAC LD audio object type and whose application domain is two-way communications in which low delay is of paramount importance. The intent is to have simple implementations, so that mixdown elements and the LTP tool are not permitted, and epconfig shall be 0, so that there is no error correction information in the bitstream. It was the consensus of the Audio Subgroup to support this profile proposal. In later discussions the profile was renamed to “Low Delay AAC Profile.”Sang-Wook Kim, Samsung, presented

12391Do-Hyung KimSang-Wook KimMiyoung Kim

Proposed Integration of BSAC and SBR

This contribution showed how to use SBR in combination with BSAC, particularly considering the implication of BSAC scalability. There was considerable support for this proposal, and the Audio Subgroup agreed to convert the contribution into a Working Draft on Integration of BSAC and SBR.

153

Sang-Wook Kim, Samsung, presented

12393Sang-Wook KimMiyoung KimDo-Hyung Kim

Proposed Study on 14496-3:2005/DCOR1:200 5

This contribution proposes changes to align the text and the reference software. Corrections consist of a replacing a probability codebook and revised syntax for decoding. It was the consensus of the Audio Subgroup to incorporate the proposal into a Study on DCOR 1 output document.

4.5.6 MPEG-4 Conformance

Andreas Schneider, Coding Technologies, presented12346 Andreas Schneider proposed study on Parametric Stereo Conformance

The Audio Subgroup agreed to turn this contribution into a Study on Parametric Stereo Conformance document.The contribution was presented:

12392Miyoung KimSang-Wook KimDo-Hyung Kim

Proposed New Audio Conformance Bitstreams for BSAC

This proposes new bitstreams. The Audio Subgroup agrees to put it in Study on Parametric Stereo Conformance document.

4.5.7 Symbolic Music Representation – Giorgio ZoiaWork of the task group can be roughly summarized in three points: core experiments, working draft, and Systems amendment.The Core Experiment methodology has been finalized after review of the draft version document (m12223) done during the AHG meeting on Sunday. The main open issue was the definition of the number of sites and experts checking the CE proposal and the meaning of statistically significant positive (or negative) assessment. Consensus on a new definition was reached. Based on this new methodology, two Core Experiments have been started at the 73rd meeting: one on Braille and Spoken Music proposed by FNB, and another one on support of Korean music proposed by KNB.Editing of the RM0 documentation (m12279) has been carried on and a first release of the Working Draft has been issued. An important point for discussion has been about what to standardize and what to leave open to implementers. It has been decided by consensus that three chunks of information will be standardized in the SMR bitstream; the most relevant one is the XML format containing the symbolic music information; this will be completed by two smaller chunks carrying high level information about formatting music symbols for rendering and spatial-temporal synchronization information with other media types respectively. Existing MPEG technology will be used for the binarization of XML, whereas the small synch info chunk (mainly numbers) may be binarized using other existing MPEG tools or by a newly defined simple method. This will be discussed during the next mandate. All the low level detail and methods for formatting and rendering will not be standardized.Finally G. Zoia (EPFL) and P. Bellini (DSI) presented contribution m12224 “Proposed Amendment to 14496-11 for SMR Integration and Support” at the Systems Group. During discussion the issue of raster vs. vector graphics has been raised; being this a decoder dependent issue, Systems experts proposed to use a more general node than the proposed one to allow the two modes at the same time. Conclusion was that a new Part 11 amendment should be started to support this functionality inside MPEG-4. The SMR task group decided to prepare a new contribution for the 74th meeting; this contribution will take into account current and possibly future remarks to have a more mature proposal, to start a WD of a new Part 11 amendment at next meeting.

4.6 Audio closing plenary discussions

5 Meeting deliverables5.1 Press statementThe Audio part of the press statement was prepared, reviewed and approved.

154

5.2 Dispositions of CommentsThe DoC were prepared and approved.

5.3 Responses to Liaison and NB commentsThe responses to the NB comments and Liaison were prepared and approved.

5.4 Recommendations for final plenaryThe Audio recommendations were presented and approved.

5.5 Establishment of Ad-hoc GroupsThe following ad-hoc groups were established by the Audio subgroup:

No. Title MtgN7393 AHG on Audio Standards Maintenance NoN7394 AHG on MPEG-7 Audio NoN7395 AHG on MPEG Surround YesN7396 AHG on MPEG-4 Lossless Audio Coding NoN7397 AHG on Symbolic Music Representation YesN7398 AHG on Exploration of Scalable Speech and Audio Coding Yes

5.6 Approval of output documentsAll output documents, shown in Annex D, were presented in Audio plenary and were approved.

6 Future activities6.1 Schedule of future meetingsAd Hoc group meetings are indicated in Section 5.5. Unless otherwise indicated, Ad Hoc group meetings will be held at the location of the next MPEG meeting on the weekend preceding that meeting.

6.2 Agenda for next meetingThe agenda for the next MPEG meeting is shown in Annex E

6.3 All other businessThere was none.

6.4 Closing of the meeting The 73rd Audio Subgroup meeting was adjourned Friday at 12:15. The Chair thanked Heiko Purnhagen, Coding Technologies, for presenting the Audio report in closing plenary, thanked the Audio experts for their hard work during the week, and urged them to enjoy their well-deserved lunch.

155

Annex A ParticipantsFirst Name Last Name Country Affiliation

Benjelloun Abdellatib FR France Telecom R&D

Jon Boley US Univ. of Miami

Gregoire CARPENTIER FR IRCAM

Kwok Chau US NJR Corp.

Chien-Fa Chen US ITRI Inc.

Elion Chin US MoodLogic

Martin Dietz DE Coding Technologies

Bernhard Feiten DE Deutsche Telekom

Matt Fellers US Dolby

Ralf Geiger DE Fraunhofer IIS

Bernhard Grill DE Fraunhofer IIS

Matthias Gruhne DE FhG IIS AEMT

Hsueh-Ming (Ming) Hang US National Chiao Tung Univ., Taiwan

Noboru Harada JP NTT

Jürgen Herre DE Fraunhofer IIS

Haibin Huang SG I2R

Inseon Jang KR ETRI

Yutaka Kamamoto JP NTT/The Univ. of Tokyo

Junghoe Kim KR Samsung AIT

Mi Young Kim KR Samsung

Minsoo Kim KR Pixtree

Sang-Wook Kim KR Samsung

SeungHynn Kim KR LG Electronics

Kristofer Kjörling SE Coding Technologies

Tilman Liebchen DE TU Berlin

Xiao Lin SG I2R

Han Gil Moon KR Seoul National Univ.

Takehiro Moriya JP NTT

Sua Hong Neo SG Panasonic

Paolo Nesi IT UNIFI DSI

156

Toshiyuki Nomura JP NEC

Takeshi Norimatsu JP Panasonic

Eunmi Oh KR Samsung

Hyen-O Oh KR LG Electronics

Werner Oomen NL Philips

Hee-Suk Pang KR LG Electronics

Sunghee Park KR Pixtree

Pierrick Philippe FR France Telecom R&D

Heiko Purnhagen SE Coding Technologies

Schuyler Quackenbush US ARL

Susanto Rahardja SG I2R

Jean Bernard Rault FR France Telecom R&D

Jonas Rdn SE Coding Technologies

Yuriy Reznik US Qualcomm

Andreas Schneider DE Coding Technologies

Jeongil Seo KR ETRI

Ralph Sperschneider DE Fraunhofer IIS

Herve Taddei DE Siemens AG

Yoshiaki Takagi JP Panasonic

Anisse Taleb SE Ericsson AB

Naoya Tanaka JP Panasonic

Yasuhiro Toguri JP Sony

David Virette FR France Telecom R&D

Ingo Wolf DE Deutsche Telekom

Dai Yang US Vidiator Technology

Rongshan Yu SG I2R

Steve Zhoa US ITRI Inc.

Giorgio Zoia CH EPFL

157

Annex B Audio Contributions and ScheduleJoint MeetingsGroups What Where Day TimeReq, Audio, MDS, Systems

Protected Music MAF (12197, 12289, 12310, 12335, 12360)

Req Tue 1600-1800

Req, Audio MPEG-4 Profile (12372) Audio Wed 1700-1730Req, Audio, MDS, Systems

Protected Music MAF Req Thu 1500-1600

Joint Meeting on Protected MAF Submitted to

12360 Stefan KraegelohHarald Fuchs

Proposed Protected Music Player MAF

Requirements

12197 HendryMunchurl Kim

Protecting and Governing Music MAF Player Format based Contents by using MPEG-21 IPMP

Audio

12289Zhongyang HuangShengmei ShenTakafumi Ueno

IPMP Protection for MPEG-A Music Player

Audio

12310

Jongheum KimSungphil HeoWeongeun OhHaekwang Kim

IPMP metadata protection for super-distribution MAF

Requirements

12335Zvi LifshitzCyril ConcolatoTIRAMISU consortium

Proposal for free-distribution MAFRequirements

Joint Meeting with Req on: Audio Profiles; MAF

Submitted to

12372 Ralph SperschneiderManfred Lutzky

Proposed new profile for high-quality communication

Audio

12396 Jeongyeon LimMunchurl Kim

Synchronization of Multiple JPEG data to MP3 tracks in Music MAF Player Format

Audio

Audio Schedule

Time Agenda ItemNumber Source Title Complete

Sunday

0900-1300 Spatial Audio Coding Room 14B/3

12226 Werner OomenFrancois Myburg

Additional subjective test results on High quality operation modes of spatial audio coding RM0.

1 X

12368

A. HoelzerW. OomenK. KjörlingJ. Herre

Additional subjective test results on R/D scalability of Spatial Audio Coding RM0 2 X

12267

Chong Kok Seng Takeshi Norimatsu Yoshiaki Takagi Neo Sua Hong

Request for formal verification test of the Low Complexity mode of Spatial Audio Coding RM0 and backward compatibility with LP-SBR

3 X

12320

Kristofer KjörlingWerner OomenFrancois MyburgJürgen Herre

Proposed corrections and clarifications to the WD on Spatial Audio Coding 4 X

12361 Kristofer Kjörling Proposal for extended signalling in Spatial Audio 5 X

12350David ViretteManuel BriandPierrick Philippe

Comments on the Spatial Audio Coding RM0 and Working Draft 6 X

158

Time Agenda ItemNumber Source Title Complete

12265

In Yong ChoiSang Bae ChonJeongil SeoInseon JangKoeng-Mo SungKyeongok Kang

Objective Measurement of Total Auditory Quality of Spatial Audio Coding 7 X

12227 Jeroen BreebaartChristof Faller

Core experiment on non-guided spatial audio coding 8 X

1400-1700 Exploration on Scalable Audio and Speech Coding

Room 14B/3

12262 Jin LiJames D. Johnston

A Progressive to Lossless Embedded Audio Coder (PLEAC) X

12253 Kevin M. Short KOZ scalable audio compression X

12311 Anisse TalebScalable Speech and Audio Coding based on the 3GPP AMR-WB+ standard: Response to the CfI on scalable speech and audio coding

X

12302 Eunmi OhYang-Seock Seo

Response to CfI on scalable speech and audio coding: The need for a new coding paradigm and its application scenarios

X

1400-1900 Symbolic Music Representation

Room 14B/204

12223Giorgio ZoiaPaolo NesiPierfrancesco Bellini

Draft Core Experiment methodology for MPEG-4 Symbolic Music Representation X

12236

Hyoung-Joong KimYong-Soo ChoiYong-Ju ChoJae-Gon Kim

The Extended Symbolic Music Representation for Korean Traditional Music. X

12279Pierfrancesco BelliniPaolo NesiMaurizio Campanai

MPEG-4 SMR RM0 documentation for SMR WD text X

Monday0900-1300 MPEG Plenary1300-1400 Lunch1400-1800 Audio Plenary 

Opening of the meeting

Administrative mattersApproval of agendaApproval of 72nd MPEG meeting report

X

12254 Schuyler Quackenbush 72nd MPEG Audio Subgroup Report X

Communications from the Chair- Review of Sun Chairs meeting- Collection of patent statements

X

Allocation of contributions to agenda and schedule

X

Joint meetings XReview of AhG reports

12124 R. Sperschneider AHG on Audio Standards Maintenance X12125 M. Gruhne AHG on MPEG-7 Audio X12126 S. Quackenbush AHG on Spatial Audio Coding X12127 Tilman Liebchen AHG on MPEG-4 Lossless Audio Coding X

12128 Paolo NesiGiorgio Zoia AHG on Symbolic Music Representation X

12129 S. Quackenbush AHG On Exploration of Scalable Audio and Speech Codi ng X

Task groups and mandates XNational body comments

12263 A. G. Tescher for USNB USNB Contribution: MPEG-4 ALS issues X

12264 A. G. Tescher for USNB USNB Contribution: Complexity issues regarding Audio tools

X

12269 CJ Lee for SGNB SGNB comment on request to extend the work schedule for Spatial Audio Coding

X

159

Time Agenda ItemNumber Source Title Complete

12271 Kyuheon Kim Korean NB's Comment on Spatial Audio Coding

X

12348 Pierrick Philippe for FNB FNB Contribution: Spatial Audio Coding Activity issues X

12349 Pierrick Philippe for FNB FNB Contribution : Comments on N7016 (signaling and decoding behavior) X

12394 Kyuheon Kim for KNB Korean NB Contribution: Public MPEG Software issue s

X

Liaison documents

12150 ITU-R SG 6 via SC 29 Secretariat Liaison Statement from ITU-R SG 6/WP 6A X

12151 ITU-R SG 6 via SC 29 Secretariat Liaison Statement from ITU-R SG 6/WP 6A X

Ballot responses and related documents

12156 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-3:2001/FPDAM 4 [SC 29 N 6748] (ALS)

X

12166 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-3:2001/FPDAM 5 [SC 29 N 6773] (SLS)

X

12155 SC 29 SecretariatSummary of Voting on ISO/IEC 15938-4:2002/PDAM 2 [SC 29 N 6747] (MPEG-7 Audio)

X

1400-1600Plenary discussions

1800- HOD Meeting

Tuesday0900 Plenary discussions

S QuackenbushFeature Extraction plugin API: http://feapi.sf.net

X

12319Kristofer KjörlingWerner OomenJürgen Herre

MPEG Surround, a proposed new name for the Spatial Audio Coding tool X

12347 Schuyler Quackenbush Proposed Tutorial on MPEG Surround Audio Coding X

12223Giorgio ZoiaPaolo NesiPierfrancesco Bellini

Draft Core Experiment methodology for MPEG-4 Symbolic Music Representation X

1000 MPEG-7

12208 Gregoire CarpentierJerome barthelemy Core experiment on weighted scales X

12241 Holger CrysandtMarkus Beermann

Cross check results of the EnhancedAudioSignature X

12367 Matthias Gruhne CE Crosscheck on Weighted Scale Type X

1100 Spatial Audio

12266

Jeongil SeoSeungkwon BeackIn Young ChoiInseon JangKyeongok KangJinwoo HongMinsoo Hahn

An Efficient Quantization Method for CLD X

12268

Hee-Suk PangDong Soo KimJaehyun LimHYEN-O OH

Subband domain energy compensation for spatial audio coding X

1300-1400 Lunch

12225Werner OomenKristofer KjörlingJuergen Herre

Request for spatial audio coding profiles X

Music Player MAF

160

Time Agenda ItemNumber Source Title Complete

12396 Jeongyeon LimMunchurl Kim

Synchronization of Multiple JPEG data to MP3 tracks in Music MAF Player Format

X

1600-1800 Joint Meeting on Protected MAF (at Req)

12360 Stefan KraegelohHarald Fuchs Proposed Protected Music Player MAF X Req

12197 HendryMunchurl Kim

Protecting and Governing Music MAF Player Format based Contents by using MPEG-21 IPMP

X Audio

12289Zhongyang HuangShengmei ShenTakafumi Ueno

IPMP Protection for MPEG-A Music PlayerX Audio

12310

Jongheum KimSungphil HeoWeongeun OhHaekwang Kim

IPMP metadata protection for super-distribution MAF

X Req

1800-1900 Liaison 1900- Chairs Meeting

Wednesday0900-1100 MPEG Plenary

1100 MPEG-4Discussion of USNB ballot comment on PS signalling

X

1300-1400 Lunch

1400 MPEG-4

12372 Ralph SperschneiderManfred Lutzky

Proposed new profile for high-quality communication X

12391Do-Hyung KimSang-Wook KimMiyoung Kim

Proposed Integration of BSAC and SBRX

12393Sang-Wook KimMiyoung KimDo-Hyung Kim

Proposed Study on 14496-3:2005/DCOR1:200 5

X

MPEG-4 Conformance

12346 Andreas Schneider proposed study on Parametric Stereo Conformance

X

12392Miyoung KimSang-Wook KimDo-Hyung Kim

Proposed New Audio Conformance Bitstreams for BSAC

X

1600 ALS/SLS Performance and Verification

12327 Tilman Liebchen Additional Information on the Performance of ALS

X

12380

Rongshan Yu Ralf Geiger Xiao Lin Susanto Rahardja

Update on Performance and Complexity of Lossless Audio Coding Architectures

X

12381

Ralf Geiger Markus Schmidt Jürgen Herre Thomas Sporer

Proposed Verification Test Methods for MPEG-4 SLS

X

1700 - 1730 Joint meeting with Req at Audio

12372 Ralph SperschneiderManfred Lutzky

Proposed new profile for high-quality communication X

1745-2200 Social

Thursday

Approval of documentsMPEG-7 items

161

Time Agenda ItemNumber Source Title Complete

SLS

12337Rongshan Yu Xiao Lin Susanto Rahardja

Integration of MPEG-4 SLS with MPEG-4 scalable audio coding tools

X

12379

Rongshan YuRalf GeigerXiao LinSusanto Rahardja

Proposed Text of ISO/IEC 14496-3:2001/FDAM 5, Scalable Lossless Coding (SLS)

X

ALS

12244Yutaka KamamotoNoboru HaradaTakehiro Moriya

Proposed corrections to avoid inconsistency in study on FPDAM4 ALS

X

12326 Tilman Liebchen Proposed Corrections and Modifications of Study on FPDAM4 (ALS)

X

12336 Haibin Huang Wee Boon Choo

Proposed Corrections to Study on MPEG-4 ALS FPDA M4

X

12325 Tilman Liebchen

Proposed Text of ISO/IEC 14496-3:2001/FDAM 4, Audio Lossless Coding (ALS), new audio profiles and BSAC extensions

X

1300-1400 Lunch

1500 - 1400 Joint meeting on Protection MAF at Req

12396 Jeongyeon LimMunchurl Kim

Synchronization of Multiple JPEG data to MP3 tracks in Music MAF Player Format

Audio

1600-1800 Audio PlenaryRecommendations for final plenary

X

Establishment of new Ad-hoc groups

X

Dispositions of commentsResponses to NB comments XLiaison statements XPress statementApproval of output documentsStatus for Chairs Meeting X

1800- Chairs Meeting

Friday

0900-1200 Audio Plenary

0900 MAF mantates at ReqSlide show MAFReport on Chairs meeting and outline of plan for the dayApproval of output documentsReview of Audio presentationAgenda for next meetingA.O.B.Closing of the Audio meeting

1200-1400 Lunch

1400- MPEG Plenary

162

Annex C Task Groups1. MPEG-A Issues

Chair: Schuyler QuackenbushMandates:1.1. Review contributions1.2. Draft output documents to reflect consensus of WG11

2. Spatial Audio CodingChair: Schuyler QuackenbushMandates:2.1. Review contributions2.2. Status of proposed CEs2.3. Discuss timeline for standardization2.4. Discuss AMD or new part of MPEG-4 or new Std2.5. Draft workplan for Spatial Audio Coding

3. Exploration of Scalable Speech and Audio CodingChair: Schuyler QuackenbushMandates:3.1. Review CfI contributions3.2. Recommend next steps

4. Symbolic Music RepresentationChair: P. Nesi, G. ZoiaMandates:4.1. Review contributions4.2. Review CE process4.3. Discuss structure of specification4.4. Discuss timeline of standardization4.5. Prepare workplan for CEs

5. MPEG-7Chair: Matthias GruhneMandates:5.1. Review contributions5.2. Produce DoC of PDAM and text of FPDAM5.3. Produce workplan

6. Lossless CodingChair: Tilman LiebchenMandates:6.1. Review contributions6.2. Discuss ballot comments and their responses6.3. Produce DoC of FPDAM and text of FDAM for ALS and SLS6.4. Produce Report on Performance and Complexity of ALS and SLS6.5. Discuss Verification of performance of ALS and SLS

7. MPEG-4 and MPEG-4 ConformanceChair: Ralph SperschneiderMandates:7.1. Review contributions7.2. Produce Study on 14496-3:2005/DCOR1:20057.3. Produce other Study on documents7.4. Study on ISO/IEC 14496-3:200X/DCOR 1

Standards under Development:Name Description Progress to14496-3:2005/DCOR 2:2005 DCOR2 Parametric Audio Coding DCOR14496-3:2001/AMD 4 ALS FDAM14496-3:2001/AMD 5 SLS FDAM14496-X Spatial Audio Coding / MPEG Surround CD? 14496-4:2001/DCOR X:200X Proposed Audio Corrections to ISO/IEC 14496-4:2004

14496-4:2001/AMD XX Parametric coding for high quality audio (INTEGRATION) PDAM15938-4:2002/AMD 2 Audio High-Level Description Extensions FPDAM

Open BallotsBallot No: ISO/IEC 14496-3:200X/DCOR 1

Doc No: SC 29 N 6754 [SC 29/WG 11 N 7130]

Information technology -- Coding of audio-visual objects -- Part 3: Audio, TECHNICAL CORRIGENDUM 1 2005-08-25

Ballot No: ISO/IEC 14496-4:2004/FPDAM 11

Doc No: SC 29 N 6681 [SC 29/WG 11 N 7149]

Information technology -- Coding of audio-visual objects -- Part 4: Conformance testing, AMENDMENT 11: Parametric stereo conformance 2005-08-26

163

Annex D Output Documents

No. Title TBP Available14496-3 MPEG-4 Audio

7363 DoC of 14496-3:2001/FPDAM 4, Audio Lossless Coding (ALS), new audio profiles and BSAC extensions

No 05-07-29

7364 Text of 14496-3:2001/FDAM 4, Audio Lossless Coding (ALS), new audio profiles and BSAC extensions

No 05-08-12

7365 DoC of 14496-3:2001/FPDAM 5, Scalable Lossless Coding (SLS) No 05-07-297366 Text of 14496-3:2001/FDAM 5, Scalable Lossless Coding (SLS) No 05-08-127367 Request for Amendment 14496-3:2005/AMD 1 No 05-07-297368 Text of 14496-3:2005/PDAM 1, Low Delay AAC Profile Yes 05-07-297369 Study on 14496-3:2005/DCOR 1, BSAC and LATM No 05-07-297370 Text of 14496-3:2005/DCOR 2, SSC No 05-07-29

7371 Draft Performance and Complexity of MPEG Lossless Audio Coding Standards

No 05-07-29

7372 Workplan for Lossless Coding No 05-07-297373 WD on Integration of BSAC and SBR No 05-08-127374 Workplan on BSAC multichannel reference software No 05/07/297375 Core Experiment Methodology for Symbolic Music Representation No 05/07/297376 Workplan for SMR Core Experiment on Traditional Korean Music No 05/07/297377 Workplan for SMR Core Experiment on Braille and Spoken Music No 05/07/297378 Text of Working Draft for Symbolic Music Representation (SMR) No 05/07/297379 Symbolic Music Representation in MPEG, an overview No 05/07/29No. Title TBP Available

14496-4 MPEG-4 Conformance7380 Study on 14496-4:2004/FPDAM11, Parametric Stereo Conformance No 05-07-297381 Proposed correction to Audio part of 14496-4:2004 No 05-07-297382 WD 2 of conformance of MPEG-1/2 Audio in MPEG-4 No 05-07-297383 Status of MPEG-4 Audio Conformance No 05-07-29No. Title TBP Available

14496-5 MPEG-4 Reference Software7462 Request for ISO/IEC 14496-5:200x AMD10 Parametric Coding for

High Quality Audio Reference SoftwareNo 05/07/29

7463 Text of ISO/IEC 14496-5:200x PDAM10 SSC, DST, ALS, SLS Reference Software

No 05/08/29

No. Title TBP Available15938-4 MPEG-7 Audio

7384 DoC on 15938-4:2002/PDAM 4 (Audio High-Level Description Extensions)

No 05-07-29

7385 Text of 15938-4:2002/FPAM 4 (Audio High-Level Description Extensions)

Yes 05-08-12

7386 Workplan for MPEG-7 Audio No 05-07-29No. Title TBP Available

23003-1 MPEG Audio Technologies: MPEG Surround7518 Proposal for New Work Item “MPEG Audio Technologies” No 05/07/297387 WD 2 for MPEG Surround No 05-07-297388 Performance of MPEG Surround RM Technology No 05-07-29

164

7389 Workplan for MPEG Surround No 05-07-297390 Tutorial on MPEG Surround No 05-07-297391 Study on Profiles for MPEG Surround No 05-07-29

No. Title TBP AvailableExploration of Scalable Speech and Audio Coding

7392 Brief Report on Call for Information on Scalable Speech and Audio Coding

No 05-07-29

165

Annex E Agenda for the 74th MPEG Audio Meeting

1.1. MPEG-4 Audio1.2. Lossless Coding1.3. MPEG Surround1.4. MPEG-7 Audio1.5. Symbolic Music Notation1.6. Exploration2. Discussion of unallocated contributions3. Meeting deliverables3.1. Press statement3.2. Dispositions of comments3.3. Responses to NB comments3.4. Responses to Liaison statements3.5. Recommendations for final plenary3.6. Establishment of new Ad-hoc groups3.7. Approval of output documents4. Future activities4.1. Agenda for next meeting5. A.O.B.6. Closing of the meeting

Agenda Item1. Opening of the meeting2. Administrative matters2.1. Approval of agenda2.2. Approval of 73rd meeting report2.3. Communications from the Chair2.4. Allocation of contributions2.5. Joint meetings2.6. Review of AhG reports2.7. Review of task groups and mandates2.8. Received national body comments and liaison matters2.9. Plenary issues3. Task group activities3.1. MPEG Maintenance

166

Annex 10SNHC

1 Opening of the Meeting

1.1 Approval of the agenda

1.2 Goals for the weekThe goals of this week are:

Review on-going AFX CE and explorations Review on-going GFX and issue FCD document Review the status of Conformance and Reference Software for Part 16 Amd. 1 Review the SNHC profiles Review and refine the SNHC Business Plan

The output documents related to SNHC are:

No. Title Editor14496-4 MPEG-4 Conformance WD 2.0 of ISO/IEC 14496-4:200x/ AMD12 (Morphing & Textures)

Jeong-Hwan Ahn

No. Title Editor14496-5 MPEG-4 Reference Software WD 2.0 of ISO/IEC 14496-5:200x/AMD9 (Morphing & Textures) Francisco Morán

No. Title Editor14496-11 MPEG-4 Scene Description and Application EngineText of ISO/IEC 14496-11:200x/COR4 Keun Ho Kim

No. Title Editor14496-16 MPEG-4 Animation Framework eXtension (AFX)DoC on ISO/IEC 14496-16/FPDAM1 Jeong-Hwan AhnText of ISO/IEC 14496-16/FDAM1 (Morphing & Textures) Marius PredaWD 2.0 of 14496-16 2nd edition Marius PredaWD 2.0 of 14496-16/AMD2 (Geometry & Shadow) Patrick GioiaText of ISO/IEC 14496-16:2004/COR2 Keun Ho KimText of ISO/IEC 14496-16:2004/DCOR3 Alexandru SalomeAFX Core Experiments Description Marius PredaMPEG-4 Animation Framework eXtension (AFX) VM 18.0 Marius PredaSNHC FAQ 14.0 Marcos AvilesOverview of FBA (Face and Body Animation) Marius PredaOverview of 3DMC (3D Mesh Coding) Euee S. JangOverview of AFX (Animation Framework eXtension) Francisco Morán

No. Title Editor14496-21 MPEG-4 MPEG-J Graphics Framework eXtension (GFX)DoC on ISO/IEC 14496-21/CD Vishy Swaminathan

167

Text of ISO/IEC 14496-21/FCD Mikael Bourges-Sevenier

MPEG-J GFX white paper Mikael Bourges-Sevenier

MPEG-J white paper Vishy Swaminathan

1.3 Standards from SNHCIn red, status reached at this meeting. In yellow, status reached at next meeting. Projects that reached International Standard status have been removed. The name of 14496-16/Amd1 has been changed to “Morphing and Textures”. The name of Amd2 of Part 16 will be “Geometry and Shadow”.

Std

Pt Edit. Project Description CfP WD CDPDAMDCOR

FCDFPDAM

FDISFDAMCOR

4 11 200x Cor.4 (Node coding table) 05/04 05/074 16 200x Amd.1 Morphing and

Textures03/07 04/07 05/01 05/07

4 16 200x Cor.2 (extrinsic camera parameters)

05/04 05/07

4 16 200x Cor.3 (Meshgrid updates) 05/07 06/014 21 200x MPEG-J GFX 03/12 04/07 04/10 05/07 06/01

168

1.4 ScheduleMonday Tuesday Wednesday Thursday Friday

:00:30:00:30:00:30:00:30:00:30:00:30:00:30:00:30:00:30:00:30:00:30:00:30:00:30:00:30:00:30

19

20

21

22

15

16

17

18

11

12

13

14

Time

08

09

10

MPEG Plenary

Lunch break

MPEG Plenary

MPEG Plenary

SNHC Plenary

Liaison

ChairsChairsSocial event

AhG set upDocument number attributionResolutions

Lunch break Lunch break

12291 w/Req.

Output documents review

SNHC PlenarySNHC BP

124031237512405

12291122881230012377

GFX status review

12216

12329

GFXB/O

1237112397Conformance/RSW

SDK Development

GFXB/O

12291 w/Req.

SNHC BPdiscussion

RSW w/ Integration

GFXB/O

GFXB/O

Lunch breakLunch break

169

1.5 Allocation of joint meetingsSub-Groups Monday Tuesday Wednesday Thursday Friday

SystemsRequirements 14:00 12:30

Integration 14:00

1.6 Rooms allocationSNHC 15/402 (Breakout for GFX 15/401)Systems 14A/4Requirements 14B/3Video 14B/2Chairs/HoD/Liaisons 14B/B

1.7 Allocation of contributions

N° Title Schedule ActivityD1 Monday D1

MPEG Plenary D1 09:00~13:30 MPEG Gen.

12130 Report of AhG on AFX documents, CEs and softwareMarius PredaJeong-Hwan AhnFrancisco Morán

12131 Report of AhG on MPEG-J GFX

Mikaël Bourges-SévenierVishy SwaminathanItaru Kaneko

Lunch Break D1 13:30~14:30

SNHC D1 14:30~16:00 SNHC Gen.

Roll call, Agenda, Goals, Web site, FAQ, etc. Mahnjin HanSNHC D1 Ref S/W

170

N° Title Schedule Activity16:30~18:00

12403 Portable Music player with 3D animation using ISO/IEC 14496-21 CD (MPEG-J GFX)

Callow (HI Corporation)Kaneko (Tokyo Polytechnic University)Sone (Tokyo Polytechnic University)

16:30 GFX

12375 MPEG-4 SDK : beyond MPEG-4 Reference Software Marius Preda Francoise Preteux 17:00 RSW

12405 Conformance bitstreams for the AFX 3D profile Raphaèle BalterPatrick Gioia 17:30 Profile

D2 Tuesday D2

GFX Breakout D2 09:00~17:00 GFX

SNHC D2 09:00~12:30 AFX

12291 Conformance Test of Simple 3D Compression Profiles and Levels

Jeong-Hwan AhnMahnjin HanSeong Won Ryuon behalf of M3DSF

09:00 Profile

12288 Results on Core Experiment of 3D Compression Stream Design

Jeong-Hwan AhnMahnjin HanSeong Won Ryuon behalf of M3DSF

09:30 CE4

12300 Preliminary Results on EE1 for D3DMC

Matthias KautznerKarsten Mueller Aljoscha Smolic Thomas Wiegand

10:00 EE1

12377 Updates to MeshGrid A. Salomie 10:30 AFX Updates

171

N° Title Schedule ActivityR. DeklerckD. CostinA. MarkovaA. MunteanuP. Schelkens

Lunch Break D2 12:30~14:00

SNHC+Requirements in SNHC D2 14:00~15:00 AFX Profiles

12405, 12291 SNHC Profiles

SNHC D2 15:00~17:00 AFX

12216 Status Report of 3DMC revision for AFX amd. 2

Eun-Young ChangSunyoung LeeDaiyong KimByeongwook MinNamho HurSoo In LeeEuee S. Jang

15:00 CE3

12329 Spatial and SNR scalable coding of WSSs Marcos AvilésFrancisco Morán 16:00 AFX

SNHC GFX mid report D2 17:00~18:00 AFX

12237 AhG on GFX, Tokyo meeting report Mikael Bourges-Sevenier 17:00 AhG report

12238 Draft updated ISO/IEC 14496-21 CD studyMikael Bourges-SevenierVishy Swaminathan

GFX

Review breakout statusD3 Wednesday D3

172

N° Title Schedule Activity

MPEG Plenary D3 09:00~11:30 MPEG Gen.

SNHC Plenary D3 11:30~12:30 SNHC Gen

12157 Summary of Voting on ISO/IEC 14496-16:2004/FPDAM 1 [SC 29 N 6749] SC 29 Secretariat Voting results

12186 Summary of Voting on ISO/IEC 14496-11:200X/DCOR 4 SC 29 Secretariat Voting results12187 Summary of Voting on ISO/IEC 14496-16:2004/DCOR 2 SC 29 Secretariat Voting results

Discussion on CE1 & CE2SNHC Business Plan presentation

Lunch Break D3 12:30~14:00

GFX Breakout D3 14:00~17:00 GFX

SNHC D3 14:00~15:00 SNHC Gen

12397 Conformance test for the update of Depth Image-based Representation

Gyeong Ja JangShinjun LeeKeun Ho KimMahnjin HanJeong-Hwan Ahn

14:00 Conformance

12371 A First Draft for Fractalish Nature Description SchemesAlberto CastilloFrancisco MoránJosé M. Martínez

14:30 AFX

SNHC discussion D3 15:00~18:00 SNHC Gen

Conformance plan and RSWSDK development discussion

D4 Thursday D4

SNHC discussion D4 10:00~12:30 General

173

N° Title Schedule ActivitySNHC Business Plan discussion Mahnjin Han et. Al.

SNHC+Requirements in SNHC D4 12:30~13:00 AFX Profiles

SNHC Profiles

Lunch Break D4 13:00~14:00

SNHC+Integration in SNHC D4 14:00~15:00 RSW

Reference software & conformance discussion

SNHC D4 15:00~18:00 SNHC Gen

Output documents reviewD5 Friday D5

SNHC D5 09:00~13:00 SNHC Gen.

Document number attribution, AhGs and resolutions

Lunch Break D5 13:00~14:00

MPEG Plenary D5 14:00~22:00 MPEG Gen.

1.8 Attendance list .Name Country Company e-mailMahnjin Han Korea Samsung AIT mjhan AT samsung DOT comMarius Preda France INT marius DOT preda AT int-evry DOT frFrancisco Morán Spain UPM fmb AT gti DOT ssr DOT upm DOT

esMikael USA Mindego mikael AT mindego DOT com

174

Bourges-SevenierSeong Won Ryu Korea ETRI ryusw AT etri DOT re DOT krEun-Young Chang

Korea ETRI eychang AT etri DOT re DOT kr

Patrick Gioia France France Telecom R&D

patrick DOT gioia AT rd DOT francetelecom DOT com

Mark Callow Japan HI Corporation callow_mark AT hicorp DOT co DOT cp

Thomas Di Giacomo

Switzerland MIRALab, U. of Geneva

thomas AT miralab DOT unige DOT ch

Karsten Mueller Germany FHG-HHI kmuller AT fraunhofer DOT hhi DOT de

Euee S. Jang Korea Hanyang Univ. esjang AT ihanyang DOT ac DOT krSunyoung Lee Korea Hanyang Univ. sunnykr AT ihanyang DOT ac DOT krJeong-Hwan Ahn Korea Samsung AIT jeonghwan DOT ahn AT samsung

DOT com Marcos Avilés Spain UPM mar AT gti DOT ssr DOT upm DOT

es

175

2 General issues

2.1 Web siteThe new official SNHC Web site has been opened at http://www.mpeg-snhc.orgThe group thanks Samsung AIT for making and hosting this official SNHC web site.

2.2 SNHC FAQThe first time attendant of SNHC subgroup was asked to review the current FAQ and give feedback. A couple of questions were raised, but only one item related to decoder & content complexity remains as follows.

3.7. Does the SNHC group also address technologies for applications targeted at limited resources terminals, such as PDAs or mobile phones?AFX was not conceived with any specific kind of terminal in mind, but the general MPEG profile and level mechanism can be used to limit the complexity of a particular content to make it suitable for limited resources terminals.

2.3 Demonstrations

2.3.1 M12403 - Portable Music player with 3D animation using ISO/IEC 14496-21 CD (MPEG-J GFX) As recent mobile devices start having mp3 players and 3D graphics capability, the combination of 3D animation added to music can be a popular function for mobile devices. This demonstration shows that MPEG-J GFX is a well designed standard format for such application.

This demonstration uses GFX to play back combined audio, video and 3D animation. Video is used for movie texture and audio is used to animate object, color and camera by FFT analysis.It currently uses C API (Mascot Capsule) but will also be done with JAVA API.It is not yet a full GFX implementation, but it shows how GFX will be beneficial in such applicationThe content has been created with MAYA and the exporter to M3G format

2.3.2 M12375 – MPEG-4 SDK : beyond MPEG-4 Reference SoftwareUntil now, MPEG-4 Systems and SNHC tools were hard to implement for the non-MPEG-expert developers. In order to make it easier for them to use MPEG-4 tools, INT has contributed MPEG-4 SDK, which will allow usage of MPEG-4 tools without knowing the details of it. It is composed of two levels: low-level API (scene graph access, media and stream processing) and high-level API (data-based semantic access, hierarchical object processing and stream control).The current implementation of MPEG-4 SDK is focused on Animated Character Profile (Virtual Character and static object)

176

In order to validate the developed SDK, 3D Studio MAX import/export module for MPEG-4 has been implemented.

ResolutionFirst of all, SNHC thanks INT for initiating the activity for developing MPEG-4 SDK.This SDK will be made available outside the MPEG community to allow implementation of AFX application (currently only virtual character animation) without having to know the details of the MPEG-4 Systems and AFX standard.This activity will be continued for extension, optimization and bug fix.

3 AFX (14496-16) activities

3.1 Specifications updatesThe list of modifications for existing specifications handled by SNHC is as follows.

3.1.1 M12377 – Updates to MeshGridWhile developing medical application with MeshGrid, few restrictions in the current specification has been identified and solutions have been proposed.

1. The ROILast can be huge. à distribute the remainder evenly across ROIs

2. Cyclic mesh is inefficient for surfaces of volumetric data (medical, scientific visualization). à By introducing the Folded Mesh, the Cyclic will be simplified

3. Allow data with many layers of meshes à add identifier to distinguish between the different surface layers

4. Morphing with topological changes à add update_replace flag

ResolutionProposed solutions have been accepted.Issue a DCOR2 of 14496-16 for fixing the restrictions in the current specification.

3.2 Profiles

3.2.1 M12405 – Conformance bitstreams for the AFX 3D ProfileBitstreams have been generated and checked for Simple 3D Scene Graph Profile and Graphics Profile in the Profile under Consideration document.There are problems with the leveling of compression tools within Simple 3D Profiles.We need 3D Compression Profile for the compression tools in scenegraph and graphics profile.

ResolutionThe corresponding 3D Compression Profile will be proposed at the next meeting for the compression tools used in the proposed profile with clear level conditions.

177

3.2.2 M12291 – Conformance Test of Simple 3D Compression Profiles and LevelsRemoval of “16.16 FixedPoint, 8-bit and 16-bit Integer” is proposed because the data type is highly related to the rendering engine functionality.Removal of WSS from the Simple 3D Compression Profile has been proposed because the companies supporting this profile think that the view dependent application through back-channel is not needed and they are not willing to add it into their product.Bitstreams have been generated for each tool in each level and mux/demux of independent bitstream has been demonstrated.

Joint meeting with RequirementsWe need to think about the future when making the profile in order to prevent making of many profiles for similar application domain.Although the proponents of this profile do not feel the need for WSS, there is a voice saying that this tool will be needed in the future.The solution would be to keep WSS with minimal limit for the level in the current profile. This was discussed between the supporting companies and consensus was reached.

ResolutionKeep WSS with minimum functionality (base mesh only) for level 1 and level 2Other updates are accepted.Bitstreams have been verified.Remove this profile from the Profile under Consideration document and move it to 14496-16/FDAM1.

3.3 Core experiments

3.3.1 CE1 – Physics based animationNo contribution for the past 2 meetings.

ResolutionClose the CE until the related contribution is received in future meetings

3.3.2 CE2 – Multi-resolution Footprint-based RepresentationNo contribution for this meeting.

ResolutionThe activity is on-going. So continue the CE

3.3.3 CE3 – 3DMC Functionalities1. M12216 – Status Report of 3DMC revision for AFX Amd. 2Reference software has been implement with the following changes from the original 3DMC ref s/w:

Deleting forest split operation

178

Deleting stitch operation Incorporating novel texture coordinate compression Incorporating lossless or near-lossless compression (SFV) Incorporating vertex and face order compression

The changes are applied to the WD of part 16 AMD2

ResolutionContinue the CEFor more functionality to be added, we want to see the support from the industry.Software for this CE, and other CE as well, will be uploaded and updated in the DataBank of SNHC homepage.

3.3.4 CE4 – 3D Compression stream design2. M12288 – Results on Core Experiment of 3D Compression Stream DesignThis contribution shows the verification result of 3D Compression Stream proposed in Busan meeting by implementing a multiplexing tool that enables the combination of various AFX streams.Minor modifications to the bitstream syntax are proposed for better efficiency.

ResolutionRename this technology as MPEG-4 3D Graphics Stream (.m3d) and add it into 14496-16/FDAM1.

3.4 Explorations

3.4.1 E1. 3D Mesh Compression Framework3. M12300 – Preliminary Results on EE1 for D3DMCIntermediate results on improvement of D3DMC with variable bit allocation depending on the degree of motion between consecutive frames of a mesh sequence.Seems to be efficient in many cases based on the experiment so far, but we need to make complete evaluation for the rest of the test content.

ResolutionContinue the EE until next meeting.The evaluation is for L1~L4 and not the original. However, the comparison should be made with original data.Also, the bitrate in the graph should be changed to bitstream size.

3.5 Reference Software & Conformance

3.5.1 M12397 – Conformance test for the update of Depth Image-based RepresentationThis is a contribution of conformance test result for the extrinsic camera parameter, which will be applied to 14496-11/COR4, 14496-16/COR2 and 14496-16/FDAM1.The test has been performed with SimpleTexture node and PointTexture node, because changing of camera parameters for OctreeImage node is not feasible. Also, due to a bug in current implementation of MovieTexture node, camera animation with movie textures are not performed

179

ResolutionThe conformance test shows that for SimpleTexture and PointTexture node with still textures, the extrinsic camera parameter works well.

3.5.2 AFX Reference Software status check (follow-up of M11955 from Busan meeting)Following the resolution of the Busan meeting, the status of WSS implementation has been checked. However, not all problems have been solved in the current reference software. Therefore, the WSS implementation will be replaced by a new implementation.

ResolutionThe WSS implementation in the reference software will be replaced with the code provided by UPM. France Telecom will crosscheck to verify its conformance.

3.6 Additional AFX related issues

3.6.1 M12371 – A First Draft for Fractalish Nature Description SchemesThis contribution reports a preliminary work on high level description of fractalish nature shape (mountains, trees, etc.) using MPEG-7 description schemes.It is a very compact representation for randomly designed models such as mountains or trees.However, it requires complex algorithm for rendering. Therefore, it may not be suitable for mobile application.Also, if we go further into specific application, the randomness is not preferred by content providers, who want to have control over the whole scene. Therefore, trees may be ok, but the terrain is not preferred by the mobile 3D game developers.However, these comments are focused on mobile 3D game application which is by far not the only application this technology is targeting.

ResolutionNeed big pictures based on functionalities and some application scenarios to make others understand and possibly support this new work. (same resolution as last meeting for M11953)Need to find out the interest in the industry for compact, high-level description (MPEG-7 DS or not)

3.6.2 M12329 – Spatial and SNR scalable coding of WSSsThe current WSS (wavelet subdivision surface) uses SPIHT-based coders which is not appropriate if used for terminals with heterogeneous capabilities. The proposed method, called PLTW (Progressive Lower Tree Wavelet) coder shows higher compression ratios than current technique.Also, it features SNR (local) scalability and spatial (global) scalability, which is attractive especially for use by wide range of terminals in terms of processing power and bandwidth.However, we need to evaluate the reconstruction error of decoded mesh and the complexity of the proposed method.

Resolution

180

Open an Exploration Experiment for further evaluating the proposed method with the current WSS.

4 GFX (14496-21) activities

4.1 M12238 – Draft of ISO/IEC 14496-21 FCDThe specification was updated following the CD ballot comments and the AhG meeting in Tokyo, May 25-26, 2005. Besides reviewing the text for correctness, the following sections needed more attention this week:

MPEG-4 Systems interface

Terminal capabilities API

Update Java code and verify correctness of the code

We reused the MPEG-J API defined in ISO/IEC 14496-11 with minor modifications so to align this API with GFX design while still preserving MPEG-J and MPEG-4 systems access features.The terminal capabilities API, originally in MPEG-J, was redesigned and simplified. Instead of having a method per capability, we followed the System.getProperty() design for static capabilities and reused the DynamicObserver design for dynamic capabilities.From our design we successfully generated the Java source code and verify the correctness of the code against our models. We started documenting the code; this documentation being normative.The specification is now believed to be complete and the FCD can be issued.

4.2 GFX reference softwareThe GFX reference software will consist of the following modules (see Figure 1) provided by various companies:

MPEG-J GFX API

MPEG-4 Systems APIs

MMAPI

Java bindings to OpenGL ES

Mobile 3D Graphics

The software will use both Java bytecode and native code running a mobile virtual machine, or KVM, running on Microsoft Windows desktop. Third-party APIs will be provided in binary and source code will be available for the rest. Examples of MPEGlets will be developed by each participant.

181

MPEG-4 Systems

MMAPI(JSR-135)

OES (JSR-239)

MPEG-J GFXKVM

M3G (JSR-184)

Java

native

MPEGlet

Figure 1 – GFX reference software overview

5 Resolutions of SNHC

5.1 Output documents

5.1.1 The SNHC subgroup recommends to approve the following documents

No. Title TBP Available14496-1 :2004/MPEG-4 Systems

7472 Text of ISO/IEC 14496-1:2004/PDAM2 3D Compression Profile Descriptor

No 05/07/29

14496-4 MPEG-4 Conformance 7347 WD2.0 of ISO/IEC 14496-4:200x/ AMD12 Morphing &

TexturesNo 05/07/29

14496-5 MPEG-4 Reference Software 7348 WD2.0 of ISO/IEC 14496-5:200x/AMD9 Morphing &

TexturesNo 05/08/15

14496-11 MPEG-4 Scene Description and Application Engine

7473 Text of ISO/IEC 14496-11:2004/COR4 Node Coding Table (Extrinsic Camera Parameter)

No 05/08/15

14496-16 MPEG-4 Animation Framework eXtension (AFX)7399 DoC on ISO/IEC 14496-16/FPDAM1 No 05/07/297400 Text of ISO/IEC 14496-16/FDAM1 (Morphing & Textures) No 05/07/297401 WD 2.0 of 14496-16 2nd edition No 05/08/197402 WD 2.0 of 14496-16/AMD2 (Geometry & Shadow) No 05/08/197403 Text of ISO/IEC 14496-16:2004/COR2 No 05/08/157404 Text of ISO/IEC 14496-16:2004/DCOR3 No 05/08/197405 AFX Core Experiments Description No 05/07/297406 SNHC FAQ 14.0 Yes 05/07/297456 Overview of FBA (Face and Body Animation) Yes 05/07/297457 Overview of 3DMC (3D Mesh Coding) Yes 05/08/12

182

7458 Overview of AFX (Animation Framework eXtension) Yes 05/08/1514496-21 MPEG-4 MPEG-J Graphics Framework eXtension (GFX)

7407 DoC on ISO/IEC 14496-21/CD No 05/07/297408 Text of ISO/IEC 14496-21/FCD Yes 05/08/157409 MPEG-J GFX white paper Yes 05/07/297410 MPEG-J white paper Yes 05/07/29

5.2 ResolutionsMPEG-4Part 16

The SNHC subgroup would like to express its gratitude to all reviewers who submitted comments on AFX FPDAM1.

The SNHC subgroup would like to thank Samsung AIT for redesigning and hosting the SNHC home page (http://www.mpeg-snhc.org).

The SNHC subgroup would like to thank INT for providing an MPEG-4 SDK for representing Virtual Character Animation and static 3D objects, and encourages interested parties to use it and provide feedback

The SNHC subgroup acknowledges the specification for the Core 3D Compression Profile designed especially for mobile 3D graphics, and encourages the mobile 3D graphics industry to use it.

Part 21 SNHC recommends that the following documents be published on the MPEG

(http://www.chiariglione.org/mpeg/) and SNHC (http://www.mpeg-snhc.org/) home pages:o ISO/IEC 14496-21 FCDo MPEG-J and MPEG-J GFX white papers

5.3 Establishment of SNHC Ad-Hoc Groups

N7411 Ad Hoc Group on AFX documents, CEs, and softwareMandate: 1. Maintain and edit SNHC documents.

2. Coordinate SNHC CE and EE activities.3. Coordinate collection of demonstration data set of SNHC tools.4. Cross-check the new implementation of the WSS decoder.5. Coordinate conformance and reference software for AMD1 (Morphing and

Textures).Chairman: Marius Preda (INT)

Co-chairs: Jeong-Hwan Ahn (Samsung AIT)Francisco Morán Burgos (UPM)

Duration: Until 74th Meeting Meetings Sunday before 74th MeetingReflector: mpeg-snhc AT gti. ssr. upm. EsSubscribe: http://www.gti.ssr.upm.es/mailman/listinfo/mpeg-snhc

183

N7412 AHG on MPEG-J Graphics Framework eXtensionsMandate: 1. Coordinate implementation of GFX reference software

2. Investigate GFX conformanceChairman: Mikaël Bourges-Sévenier (Mindego Inc.)

Co-chairs: Vishy Swaminathan (Sun Microsystems)Itaru Kaneko (Tokyo Polytechnic University)

Duration: Until 74th Meeting Meetings Sunday before 74th MeetingReflector: mpeg-j AT it-aru. ComSubscribe: Send an email to mpeg-j-request AT it-aru. com with the message “subscribe” as

the first line in the body.

6 Closing of the MeetingSee you in Nice.

184

Annex 11Integration

1 Opening of the Meeting

1.1 Allocation of contributionsMonday Plenary

12116 Yi-Shin TungChung-Neng Wang AHG on Maintenance of MPEG-4 Visual related Documents, Reference Software and Conformance

12117

L. CieplinskiA. YamadaSang-Kyun KimRobert O'Callaghan

AHG on Maintenance of MPEG-7 Visual related Documents and Reference Software

12124 R. Sperschneider AHG on Audio Standards Maintenance

12130Marius PredaJeong-Hwan AhnFrancisco Morán

AHG on AFX documents, CEs, and software

Integration Plenary on Thursday, 2pm to 4pmx Francisco Moran Burgos Maintenance of MPEG-4 Reference Software

12394 Kyuheon Kim for KNB Korean NB Contribution: Public MPEG Software issues

12163 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-4:2004/FDAM 6 [SC 29 N 6755]

12164 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 15938-7:2003/FDAM 1 [SC 29 N 6756]

12190 SC 29 Secretariat Summary of Voting on ISO/IEC CD 21000-14

1.2 List of standards under developmentStd Pt Edit. Project Description WD CD FCD FDIS

1 4 200x Cor 1 05/07 06/012 4 200x Cor 1 05/07 06/014 4 200x Amd.9 AVC FRExt Conformance 04/03 04/07 05/01 05/074 4 2004 Amd.11 Parametric Stereo Conformance 05/01 05/04 05/104 4 2004 Amd.12 Morphing and Texture conf 05/04

4 4 200x Amd.13 Parametric Coding 05/07 06/01 06/074 4 200x Cor 2 05/07 06/014 5 2004 Amd.8 AVC FRExt Ref Soft 04/07 05/01 05/074 5 200x Amd.9 Morphing and Texture RS 05/04

4 5 200x Amd.10 Parametric Coding RS 05/07 06/01 06/074 5 200x nth Ed. Reference Software 05/07

21 14 200x 1st Ed. Conformance 03/10 05/04 05/10 06/04

1.3 Latest references

Project P. Standard Issue No.MPEG-4 4 ISO/IEC 14496-4:2003 (MPEG-4 Conformance 2nd Ed.) 02/12 Awaji N5457MPEG-4 4 ISO/IEC 14496-4:2001/Amd.1 (FlexTime, Studio and Streaming

Video Profile)02/07 Klagenfurt N5083

MPEG-4 5 ISO/IEC 14496-5:2001 (Ref. Soft 2nd Ed.) 01/07 Sydney N4368MPEG-4 5 ISO/IEC 14496-5:2001/Amd.1 (FlexTime) 02/03 Jeju N4711MPEG-4 5 ISO/IEC 14496-5:2001/Amd.2 (XMT) 02/05 Fairfax N4865MPEG-7 6 ISO/IEC 15938-6:2002 (Reference Software) 01/12 Pattaya N4475MPEG-7 7 ISO/IEC 15938-7:2002 (Conformance) 02/07 Klagenfurt N4937

185

MPEG-21 8 ISO/IEC 21000-8:2005 (Reference Software) 05/04 Busan

1.4 Request documentsProj P A Description Issue No

2 2 1 Request for New Amendment: ISO/IEC 13818-2:2000/Amd 1 30614 4 1 Request for amendment 1 to ISO/IEC 14496-4 35044 4 2 Request for ISO/IEC 14496-4/Amd.2 35704 5 2 Request of ISO/IEC 14496-5/Amd.2 38644 4 3 Request for ISO/IEC 14496-4/Amd.3 38684 4 1 Request for Amendment 1 of ISO/IEC 14496-4:2001 40884 5 1 Request for Amendment 1 of ISO/IEC 14496-5:2001 40894 5 2 Request for Amendment 2 of ISO/IEC 14496-5:2001 42794 4 3 Request for Amendment 3 for ISO/IEC 14496-4 47074 5 3 Request for ISO/IEC 14496-5:2001/Amd.3 486521 8 Request for Subdivision to ISO/IEC 21000-8 49044 4 3 Request for ISO/IEC 14496-4:2002/AMD3 50854 4 4 Request for ISO/IEC 14496-4/AMD4 54424 5 4 Request for ISO/IEC 14496-5/AMD4 54444 4 5 Conformance Extensions for Error Resilience Scalable Profile 56694 5 5 Error Resilience Scalable Profile Reference Software 56702 4 1 IPMP Conformance Extensions 57992 5 1 IPMP Reference Software Extensions 58084 4 6 AVC Conformance 58154 5 6 AVC Ref Soft 58227 7 1 Conformance Extensions 58254 4 4 IPMP Conformance Extensions 58954 4 7 AFX Conformance Extensions 58964 5 4 IPMP Reference Software Extensions 58974 5 7 AFX Reference Software Extensions 589821 14 Request for Subdivision ISO/IEC 21000-14 MPEG-21 Conformance 60702 4 2 Request for Amd 2 of ISO/IEC 13818-4:200X 62334 4 8 Request for Amd 8 of ISO/IEC 14496-4:2003 62414 4 9 AVC Fidelity Range Extensions Conformance 63524 5 8 AVC Fidelity Range Extensions Reference Software 63554 4 10 Request for 14496-4:2004 Amd.10 64974 4 11 Parametric Stereo Conformance 05/01 4 4 13 Parametric Coding for High Quality Audio Conformance 05/07

2 MPEG-4 Conformance (14496-4)12163 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-4:2004/FDAM 6 [SC 29 N 6755]

The FDAM is approved unanimously.

186

3 MPEG-4 Reference Software (14496-5) Maintenance of CVS and reference software: Francisco Moran Burgos reports on his activity of maintenance of the MPEG-4 Systems reference software. This activity is going on on a best effort basis. On the licensing front, Francisco and Jean-Claude have an action to send email on the general MPEG reflector to ask help in identifying possible problems.

12394 Kyuheon Kim for KNB Korean NB Contribution: Public MPEG Software issues

The comment refers to AVC software and Parametric Audio software. The AVC software is available on the ISO site, and the audio software amendment is just entering PDAM stage.

4 MPEG-7 Conformance

12164 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 15938-7:2003/FDAM 1 [SC 29 N 6756]

The FDAM is approved unanimously.

5 MPEG-21 Conformance12190 SC 29 Secretariat Summary of Voting on ISO/IEC CD 21000-14

This will be dealt with at the next meeting during the preparation of the FCD document.

187

6 Resolutions of Integration

6.1.1The Integration and Video subgroups recommend to approve the following documents

No. Title TBP Available14496-4 MPEG-4 Conformance

7306 DoC on ISO/IEC 14496-4:200x FPDAM9 AVC FRext Conformance N 05/07/297307 Text of ISO/IEC 14496-4:200x FDAM9 AVC FRext Conformance N 05/07/29

6.1.2The Integration and Audio subgroups recommend to approve the following documents

No. Title TBP Available14496-4 MPEG-4 Conformance

7460 Request for ISO/IEC 14496-4:2004 AMD13 Parametric Coding for High Quality Audio Conformance

N 05/07/29

7461 Text of ISO/IEC 14496-4:2004 PDAM13 Parametric Coding for High Quality Audio Conformance

N 05/07/29

6.1.3The Integration and Video subgroups recommend to approve the following documents

No. Title TBP Available14496-5 MPEG-4 Reference Software

7308 DoC on ISO/IEC 14496-5:200x FPDAM8 AVC FRext Reference Software

N 05/07/29

7309 Text of ISO/IEC 14496-5:200x FDAM8 AVC FRext Reference Software

N 05/07/29

6.1.4The Integration and Audio subgroups recommend to approve the following documents

No. Title TBP Available14496-5 MPEG-4 Reference Software

7462 Request for ISO/IEC 14496-5:200x AMD10 Parametric Coding for High Quality Audio Reference Software

N 05/07/29

7463 Text of ISO/IEC 14496-5:200x PDAM10 SSC, DST, ALS, SLS Reference Software

N 05/08/29

6.1.5The Integration and SNHC subgroups recommend to approve the following documents

No. Title TBP Available14496-4 MPEG-4 Conformance

7347 WD2.0 of ISO/IEC 14496-4:200x/ AMD12 Morphing & Textures N 05/07/29

6.1.6The Integration and SNHC subgroups recommend to approve the following

188

documents

No. Title TBP Available14496-5 MPEG-4 Reference Software

7348 WD2.0 of ISO/IEC 14496-5:200x/AMD9 Morphing & Textures N 05/07/29

No. Title TBP Available21000-14 MPEG-21 Conformance

6.1.7The Integration subgroup recommends to appoint Sylvain Devillers and Thomas De Martini as editors of 21000-14, and to remove Jean-Claude Dufourd from the editors list.

189

Annex 12Test

Source: Tobias Oelbaum

Opening of the Meeting

Goals for the weekThe goals of this week are:

Viewing of results of anchor encoding for the test sequences under consideration for the CfP on Multi View Coding

Define test conditions and test method for the CfP on Multi View Coding Select test sequences for the CfP on Multi View Coding Set up timeline for testing the CfP on Multi View Coding Perform a visual quality assessment for SVC CE6 “Update Step”

Joint MeetingsThe following joint meetings were scheduled

with Video – on 3DAV CfP Multi View Coding with JVT – on SVC CE6 “Update Step”

Test Activities

CfP on Multi View CodingAs announced at the 72nd meeting new test sequences for MVC were made available. These test sequences were produced using different camera settings (number of cameras ranging from 8 to 100 with baseline distance from 5cm to 30 cm) and in different environments (indoor/outdoor). All test sequences were subject to anchor encoding to be able to finally select the test sequences and correct bit rates for the CfP.Out of 20 sequences available 8 test sequences were chosen. The selected test sequences cover a wide range of camera parameters and content. Bit rates were chosen after a viewing session of the results of the anchor encoding that has been done before and at the meeting. Selected bit rates range from 192 kBit/s up to 2Mbit/s.Camera parameters were made available for all sequences, however for some of the sequences verification for this camera parameters was missing.It was decided to use the SSIS (Single Stimulus Impairment Scale) test method to test the visual quality of the proposals. The reason for selecting this method is the wide quality range that is expected. This method has proven to deliver reliable results for coded video with a wide range of quality during the test for the CfP on Scalable Video Coding.The tests will be performed at TUM in advance to the 75th meeting starting from January 2nd 2006.If sequences from the selected test set contain two views which form a stereo view pair, additional tests will be performed. These additional tests will be done with a stereo display setup using shutter

190

glasses and a CRT monitor. However results of these additional tests will not be part of the official test report for the CfP but will be presented in an additional input document to the 75th meeting.

SVC CE6 testingIt was requested to have a subjective evaluation of the proposals to CE6. This evaluation should be done during the 73rd meeting.The test was done using the SSIS test method that was chosen for all tests for SVC so far. The test room was equipped with a DLP Projector from BenQ with a resolution of 800x600 pixels. The image was projected to the white wall, as projection on the provided screen was distorted by aliasing artefacts caused by the different resolution of the projector and the screen. Tests were performed for CIF and 4CIF cases only.Only 4 people took part in the test.As the differences between the different proposals were very small and the test conditions did not allow a thorough evaluation of subtle differences no significant differences could be found during this evaluation.Considering the experience from this evaluation the procedure of subjective evaluation of CEs at a meeting should be refined.

AVC 4:4:4 profile There is a possible need for visual tests for proposals for the replacement of the 4:4:4eren profile. However this profile and intended applications aim at a quality level above visual transparency. Reducing the bit rates to reach a quality level where differences could be seen is not a possible option, as differences between the proposals are different at different quality levels and PSNR values suggest higher gains only at very high bit rates.Despite these predictable problems the possibility of a visual quality assessment should be discussed and possible test sequences as well as test facilities should be part of this discussion.

Test Resolutions

Output Documents N7327 Call for Proposals on Multi-View Video Coding

AdHoc GroupsNo AdHoc Groups have been set up at this meeting.

191

Annex 13ISG

Source: ISG ChairEditor: Ihab Amer (UoC) and Marco Mattavelli (EPFL)

1 OverviewThe main work items of the Implementation Studies Subgroup in Poznan are:

1. The advances in the development of the MPEG-4 “Part 9 Reference Hardware Description” Phase 2 concerning the integration of the MPEG-4 Optimized Reference software Part 7 and the Reference Hardware Part 9 so as to constitute a “mixed” software hardware description of MPEG-4 video using the concept of the virtual socket.

2. The planning and extension of supported features for the integrated framework, putting together in a single application MPEG-4 Part 9 with MPEG-7 Part 7 and AVC (MPEG-4 Part-9) software.

Input contributions w.r.t. the above items are summarized according to the following table:

Contributions

M12123

“AHG on MPEG-4 Part 9 Reference Hardware Description Phase 1 and 2”

Robert TurneyMarco Mattavelli

M12292

“A SystemC Modeling Strategy to Integrate Selected IP-Blocks with the H.264 Software Reference Model”

Ihab Amer, Mohammed Sayed, Wael BadawyGraham JullienAdvanced Technology Information Processing Systems (ATIPS), Calgary, Alberta, Canada.

M12294

“Updated Hardware Blocks and Conformance Tests for the Multiple IP-Core Hardware-Accelerated Software System Framework for MPEG4”

Yifeng Qiu, Wael BadawyLaboratory for Integrated Video Systems (LIVS), University of Calgary, Alberta, Canada.

M12295

“An Integrated Virtual Socket Hardware-Accelerated Software Framework for MPEG-4”

Yifeng Qiu, Wael BadawyLaboratory for Integrated Video Systems (LIVS), University of Calgary, Alberta, Canada.

M12357

“Hardware Implementation of AVS 8x8 Inverse Integer Cosine Transform Module”

Qian Hu, QingXiao Jiang, Lu YuZhejiang University

192

2 Detailed Report

2.1 The progress in the development of the MPEG-4 “Part 9 Reference Hardware Description”

The ISG activity at the Poznan meeting has mainly been devoted to:

the review of the contributions presenting updated HDL module submissions, the review of contributions on the new proposed improvements to the API specification and

implementation improvements of the “Virtual Socket”, the update of MPEG-4 Part 2 and Part 10 module submission status, the contributions to the call for proposals for integer DCT/IDCT and the possible answers

provided by the modules submitted for Part 9

During the meeting a large portion of the time has been devoted to review input contributions. M12292 presents an approach to validate HW block within the reference software using SystemC. SystemC is a hardware design concept that enables the designer to perform early functional verification of developed hardware blocks by facilitating their integration with software in a unified platform. It provides hardware-oriented constructs within the context of C++ as a class library implemented in standard C++. In this contribution, a strategy that enables us to emulate a model of a full HW/SW H.264 encoder is presented. The latest reference software is modified by allowing selected computationally extensive modules to be optionally executed in emulated hardware. SystemC is used for hardware modeling. The strategy enables the MPEG designer to perform early functional verification and conformance analysis of the IP-blocks at the system level of abstraction.Contribution M12294 illustrates the framework for conformance testing for a hardware-accelerated software system. In this framework, generic system hardware blocks are used for testing, data communication and control signal exchange between a host computer and a configured co-processor, which is based on Annapolis Wildcard II platform. These generic blocks use hardware handshake, asynchronous or DMA data transfer techniques that provide the infrastructure for the purpose of user IP-cores that were not specifically designed for interaction with a host computer system. The user IP-core can be any hardware module designed to handle a computationally extensive task. The main point of this contribution document is to stress the concept of having a generic platform to illustrate the possibility of rapidly developing a working system for MPEG4 video applications.M12295 document presents the status of an integrated virtual socket and its conformance testing for a hardware-accelerated software system developed at the University of Calgary for the support of MPEG-4 Part-9 standard. In this framework, generic system hardware blocks are used for testing, data communication and control signal exchange between a host computer and a configured co-processor based on the concept of virtual socket. These generic blocks use four types of virtual socket memory space inside the platform to transfer data and thus provide the infrastructure for the purpose of user IP-blocks that were not specifically designed for interaction with a host computer system. The user IP-block can be any hardware module designed to handle a computationally extensive task for MPEG-4 video or AVC.

193

The important issue of this document is to stress the concept of having a generic virtual socket platform to illustrate the possibility of rapidly developing a working system for MPEG-4 video applications.Finally contribution M12357 presents the HW implementation of the of AVS Inverse Integer Cosine Transform Module (IICT) used in AVS to transform de-quantized Two-Dimension Coefficient Matrix to Two-dimension Residual Matrix. This design is based on a 8×8 block as a processing unit.

The ad-hoc group on the development of MPEG-4 Part 9 (N7274) has been re-established with updated mandates, including a specific mandate for the specification and development of the demonstration platform and the mandate for continuing the investigating the hardware reference description for DCT/IDCT as answer to the Call issued by video and JVT.

The ad-hoc schedule includes 4 telephone conferences before next meeting. Phone conferences are planned on the 25th August, 15th September 13th October at 4 p.m. CET. Tel: (from US 1-877-582-3182, from outside US 1-770-970-4161, participant code 9202060193).

2.2 Contributions and synergies with the Video Coding Tool Repository activity

A joint meeting has been held with the VCTR subgroup to explore the possible synergies and common points between Part 9 and VCTR. First, the possible collaboration between the 2 groups has been discussed. VCTR proposed of using a reference software that is being developed in VCTR and implementing the 8x8 transform on the Wildcard platform and virtual socket implemented by the ISG group. Discussions took place about the resources that ISG should devote to support VCTR proposal. ISG answered that there are the necessary resources for such a collaboration.A second issue still unresolved is the formalism and associated language and tools to be used to pass from what is currently a textual description of the tools to a true functional description that enables tool reconfigurability and simulations. ISG has not yet provided information and proposals for the formalism and for the usage of an appropriate language. Third point of discussion was the presentation of ISG to VCTR of some AVS modules. The objectives of VCTR when modules that do not belong to current existing MPEG standards, are used in the VCTR framework remain still unclear for the moment and will be resolved in future meetings..

3 ResolutionsThe above activities have led to the following resolutions and output document approval.

MPEG-4

4.7 Part 9 Reference Hardware Description related resolutions

4.7.1 The Implementation Studies subgroup recommends approval of the following documents:No. Title TBP Available

194

14496-9 MPEG-4 Reference Hardware Description N7273 Status of HDL submissions and commitments for MPEG-4 Part-9 N 05/08/10

N7275 Study of Part ISO/IEC PDTR 14496-9:2005 2nd Edition Reference Hardware Description

N 05/08/25

4.7.2 The Implementation Studies subgroup would like to thank all submitters of HDL code and contributors to the implementation of the virtual socket and encourages them to continue this fruitful effort and collaboration.

195

Annex 14Liaison

Source: Jan Bormans (Jan dot Bormans at imec dot be)

The Liaisons Group considered the following input documents:

Input Contribution Number

Title

M12146 Liaison Statement from ISMAM12147 Liaison Statement from the TV-Anytime ForumM12149 Liaison Statement from 3GPPM12150 Liaison Statement from ITU-R SG 6/WP 6AM12151 Liaison Statement from ITU-R SG 6/WP 6AM12153 IEC CDV 61966-2-4

M12165 SC 37 NP on Biometric data interchange formats -- Face identity data

M12170 Liaison Statement from DMPM12182 Liaison Statement from TVAM12210 Liaison Statement from OMAM12398 Liaison Statement from ECMA

M12407 User Needs Summary prepared by TG1 of JTC1 SWG on Accessibility

M12409 Liaison Statement from WG 1M12413 Liaison Statement from DMPM12414 SC 37 NP on 3 Dimensional Face Image Data Interchange FormatM12415 SC 37 NP on Biometric Sample QualityM12417 Liaison Statement from Khronos OpenMax

During the Liaison meeting, the observation was made that TC 42 WG18 could be a potential customer for MPEG’s MAF technology. Consequently, a Liaison Statement has been drafted to inform TC 42 WG18 of this potential complementarity.

We have also been informed of the ongoing work by TG1 of JTC1 SWG on Accessibility. It has been decided to proactively act and request the HoDs to encourage NBs to study the document produced (M12407) – see also the HoD resolutions.

Following the request of the USNB, Mike Ksar was appointed Liaison Representative to and from SC 2.

In addition to replying, when appropriate, to the incoming Liaison Statements of this meeting, it was agreed to take the initiative to send outgoing Liaison Statements to: TC 42 WG18 (on MAF) SMPTE (on OpenType® and ISO/IEC 14496-22) 3GPP (on loss resilience video experiments)

196

We also informed the following organisations of the work has been initiated for the Joint Video Team of ISO/IEC MPEG and ITU-T SG16 Q.6 (VCEG) to add three additional entries to Table E-1 of the ITU-T Rec. H.264 | ISO/IEC 14996-10 (MPEG-4 Part 10) Advanced Video Coding (AVC) specification: DVB TM-AVC ATSC S6 SCTE DVS DVD Forum WG-1 Blu-ray Disc Association TEG-2

Finally, the responses for the non-technical National Body comments as well as the updated list of organisations with which MPEG entertains liaisons were completed.

The following recommendations were issued:

The Liaison subgroup recommends the approval of the following documents:

Liaison statements

No. Title TBP AvailableGeneral

7351 Liaison Statement to ITU-R SG 6/WP 6A on ITU-R BS.1387-1 N 05/07/297352 Liaison Statement to ITU-R SG 6 WP 6A on objective evaluation

of spatial audio reproductionN 05/07/29

7353 Liaison Statement to SMPTE on OpenType® and ISO/IEC 14496-22

N 05/07/29

7354 Liaison Statement to ECMA on Universal 3D File Format N 05/07/297355 Liaison Statement to JTC1 SC 37 N 05/07/297356 Liaison Statement to 3GPP N 05/07/297357 Liaison Statement to IEC TC100 N 05/07/297358 Liaison Statement to TVAF N 05/07/297359 Liaison Statement to OMA N 05/07/297360 Liaison Statement to DMP N 05/07/297361 Liaison Statement to TC 42 WG18 N 05/07/297362 Liaison Statement to Khronos OpenMax N 05/07/297413 Liaison Statement to DVB TM-AVC N 05/07/297414 Liaison Statement to ATSC S6 N 05/07/297415 Liaison Statement to SCTE DVS N 05/07/297416 Liaison Statement to DVD Forum WG-1 N 05/07/297417 Liaison Statement to Blu-ray Disc Association TEG N 05/07/297459 Liaison to 3GPP on loss resilience video experiments N 05/07/29

Liaison documents

No. Title TBP AvailableGeneral

7349 Responses to National Body Comments N 05/07/297350 List of Organisations with which MPEG entertains liaisons (as of N 05/07/29

197

July 2005)

Nomination of Liaison Representatives

Name Organisation in Liaison To/fromMike Ksar SC 2 To/from

198