Real-Time 3D Articulated Pose Tracking using Particle Filters Interacting through Belief Propagation

Abstract

This article proposes a new statistical model for fast 3D articulated body tracking, similar to the loose-limbed model, but where inter-frame coherence is taken into account by using the previous marginal probability of each limb as prior information. Belief propagation is used to estimate the current marginal for each limb. All probability distribution are represented as sums of weighted samples. The resulting algorithm corresponds to a set of particle filters, one for each limb, where the weight of each sample, after the standard evaluation, is recalculated by taking into account the interactions between limbs. Applied to upper-body tracking in disparity and color images, the resulting algorithm estimates the body pose in quasi real-time (12Hz)