ASL Sign Recognition using SVM, Decision Trees and Neural Networks

In this project, we develop a computing system that can understand human gestures. In specific, we attempt to recognise ASL signs for the following words:

1. ABOUT https://www.signingsavvy.com/sign/ABOUT/32/1
2. AND https://www.signingsavvy.com/sign/AND/888/1
3. CAN https://www.signingsavvy.com/sign/CAN/524/1
4. COP https://www.signingsavvy.com/sign/COP/3203/1
5. DEAF https://www.signingsavvy.com/sign/DEAF/102/1
6. DECIDE https://www.signingsavvy.com/sign/DECIDE/781/1
7. FATHER https://www.signingsavvy.com/sign/FATHER/3440/1
8. FIND https://www.signingsavvy.com/sign/FIND/146/1
9. GO OUT https://www.signingsavvy.com/sign/GO+OUT
10. HEARING https://www.signingsavvy.com/sign/HEARING/3594/1

Phase 1: Data collection
Phase 2: Feature Extraction, PCA
Phase 3: Training SVM, Decision Tree and Neural Networks, User Dependent Analysis
Phase 4: User Independent Analysis

For more details, please check the specification present under each directory.