A hybrid HMM/particle filter framework for non-rigid hand motion recognition

In sign-language or gesture recognition, articulated hand motion tracking is usually the first step before achieving behaviour understanding. However the non-rigidity of the hand, complex background scenes, and occlusion, make tracking a challenging task. In this paper, we present a novel hybrid HMM/particle filter framework for simultaneously tracking and recognition of non-rigid hand motion. The novel contribution of the paper is that we unify the independent treatments of non-rigid motion and rigid motion into a single, robust Bayesian framework and demonstrate the efficacy of this method by performing successful tracking in the presence of significant occlusion clutter.