TR#379: Invariant features for 3-D gesture recognition

Lee W. Campbell, David A. Becker, Ali Azarbayejani, Aaron F. Bobick, Alex Pentland

Submitted to Second International Workshop on Face and Gesture Recognition, Killington VT, Oct. 1996

Ten different feature vectors are tested in a gesture recognition task which utilizes 3D data gathered in real-time from stereo video cameras, and HMMs for learning and recognition of gestures. Results indicate velocity features are superior to positional features, and partial rotational invariance is sufficient for good performance.