21
LEAP MOTION Shikhar Rastogi CSE 3 rd Year 1350110022

Leap Motion ppt

Embed Size (px)

Citation preview

Page 1: Leap Motion ppt

LEAP MOTION

Shikhar RastogiCSE 3rd Year1350110022

Page 2: Leap Motion ppt

Content

• Human Computer Interaction• What is Leap• Hardware and software specifiation • Classification of gestures• Behind Leap: SVM & HMM• Features• Applications• Conclusion

Page 3: Leap Motion ppt

Evolution OfHuman Computer Interaction

Effort

Tech

nica

l pe

rfor

man

ce

Mouse

Touch Screen

GestureRecognition

Page 4: Leap Motion ppt
Page 5: Leap Motion ppt

What Is The Leap?“It’s more accurate than a mouse, as reliable as a keyboard, and more sensitive than a touchscreen” (Leap Motion, 2012).

• Small rectangular gadget of size 7x3 CM.• Equipped with tiny cameras and IR sensors.• Designed to sit right in front of your system and sense movements with startling accuracy.

Page 6: Leap Motion ppt

Gestures classificatio

n

ActionGestures

Parameterised

Gestures static dynamic

global local mixe

d

global local mixe

d

Gestures

Page 7: Leap Motion ppt

H/w and S/w SpecificationSoftware:• Windows 7 and

8• Mac OSHardware:• Intel i3/i5/i7

Processor• RAM 2GB• USB 2.0 Port  

Page 8: Leap Motion ppt

Behind Leap:Michael Buchwald and David

Holz

Page 9: Leap Motion ppt

SVM: Support Vector Machine

Leap motion

controller

Pre-processing

Computing Feature set SVM Postprocessi

ngRecognised

Static gesture

Processing blocks of a static gesture recognition

Page 10: Leap Motion ppt

HMM: Hidden Markov Model

• A finite N-element set of states(graph),which are associated with probability distribution.

• Markov model is based on Markov property.

• Markov Property:-The next, future state depends on previous state and transition probability between those states.

Page 11: Leap Motion ppt

• HMM consists of (T,E, π(pi))- a. Transition probabilities stored in NxN matrix T. b. K- element observation set stored in NxK matrix E. c. N-element vector of initial state probabilities π (pi).

• Algorithms used in HMM:- a. Forward-Backward Algo. b. Viterbi Algo. c. Baum-Welch Algo.

Contd…

Page 12: Leap Motion ppt

Record videoConvert into

frames

Hand Segmentation

Feature Extraction

Pattern recognition

Perform Associated Tasks

Based on Skin colour detection

Pattern match

Processing Blocks of HMM based system

Page 13: Leap Motion ppt

Common Dynamic Gestures

NextClose Previous

Zoom -in

Rotate Anti-clockwise by

90

Zoom -out

Rotate clockwise by

90

Print

Page 14: Leap Motion ppt

Features• Using a USB connection, it creates 4-8 cubic feet of space to

interact with your computer.• It can sense movements 150 degrees lengthways and

widthways and to a depth of 120 degrees from top to bottom.• It can differentiate Chopsticks and fingers.• 200 times more accurate than existing Microsoft’s Kinect.

Page 15: Leap Motion ppt

Applications

Page 16: Leap Motion ppt

Educational Applications

• 3D modelling.• Manipulating 3D diagrams and

molecules.• Signing a digital document.

Page 17: Leap Motion ppt

Gaming

Page 18: Leap Motion ppt

Music and Video• Learning and playing instruments.• DJ-Music mixing.• Video editing.

Page 19: Leap Motion ppt

Medical• Manipulating screens• Viewing reports

Robotics• Self navigation• Mimic human movements

Page 20: Leap Motion ppt

ConclusionEasy As 1, 2, 3!

1. Plug it in

2. Load it up

3. Calibrate it

Enjoy!

Page 21: Leap Motion ppt

THANK YOU