Deformable model and HMM-based tracking, analysis and recognition of gestures and faces
Dimitrios Metaxas · 2003
In this paper we present a framework for the shape and motion estimation and recognition of faces and gestures. We first present physics-based modeling techniques for the 3D shape and motion estimation of humans based on single and multiple views as well as the integration of visual cues such as edges and optical flow. We then demonstrate that the reliable recognition of gesture and American Sign Language (ASL) in particular, requires the use of 3D tracking data, ASL phonology and modifications to the traditional use of Hidden Markov Models.