5
ABSTRACT To develop an audio signal processing algorithm that detects rales (gurgling noises that are a distinct symptom of common respiratory diseases in poultry). To derive features from the audio by calculating mel frequency cepstral coefficients (MFCCs), clustering the MFCC vectors, and examining the distribution of cluster indices over a window of time. The features are classified with a C4.5 decision tree i

Abstract

Embed Size (px)

DESCRIPTION

project abstract

Citation preview

Page 1: Abstract

ABSTRACT

To develop an audio signal processing algorithm that detects rales (gurgling noises that

are a distinct symptom of common respiratory diseases in poultry). To derive features

from the audio by calculating mel frequency cepstral coefficients (MFCCs), clustering

the MFCC vectors, and examining the distribution of cluster indices over a window of

time. The features are classified with a C4.5 decision tree

i

Page 2: Abstract

CONTENTACKNOWLEDMENT

ABSTRACT i

CONTENT ii

LIST OF FIGURES iii

1 INTRODUCTION 1

2 LITERATURE SURVEY 3

3 COMPRESSIVE SENSING 5

3.1 Fundamentals of CS 5

3.1.1 Sparsity 5

3.1.2 Incoherence 6

3.1.3 The L0, L1, and L2 Norms 6

3.2 The Underlying Matrix Problem 7

3.3 Compressible Signals 8

3.3.1 Compressive Sensing of a Signal 9

3.4 Compressive Imaging 12

4 SIGNAL RECOVERY ALGORITHMS 13

4.1 Convex Relaxation 13

4.1.1 L1 Minimization Algorithm 13

4.2 Greedy Algorithms 13

4.3 Combinatorial Algorithms 14

5 COMPRESSED SENSING MRI 15

5.1 Requirements for Application of CS to MRI 15

5.1.1 Transform Sparsity 15

5.1.2 Pseudo-Random Undersampling 16

5.1.3 Nonlinear Reconstruction 17

5.2 Compressed sensing techniques 17

5.2.1 k-t FOCUSS 18

5.2.2 k-t BLAST 18

5.2.3 k-t FOCUSS for Dynamic MRI 19

5.2.4 Bayesian Experimental Design 20

5.2.5 Modified CS 20

5.2.6 Motion Compensated Modified CS (MC-MCS) 20

5.3 Mask Design Problems 21

6 IMPLEMENTATION 23

6.1 Hardware and Software Requirements 23

ii

Page 3: Abstract

7 APPLICATIONS 24

8 ADVANTAGES AND DISADVANTAGES 25

CONCLUSION 26

REFERENCES 27

LIST OF FIGURESFig: 3.1 Random samples of the original signal generated by the “A” key on a touch-tone phone. 9

Fig: 3.2 The inverse discrete cosine transform of the signal 10

Fig: 3.3 The L1 solution to Ax = b 10

Fig: 3.4 dct(x), a signal that is nearly identical to the original 11

Fig: 3.5 The L2 solution to Ay = b leads to dct(y), a signal that bears little resemblance to the original 11

Fig: 3.6 Single pixel, compressive sensing camera 12

iii

Page 4: Abstract

Fig: 5.1 K-space trajectories Cartesian, Radian and Spiral 21

iv