A Conditional Random Field with Loop and Its Inference Algorithm

A new algorithm for human motion Recognition based on Conditional Random Fields (CRFs) and Hidden Markov Models (HMM) -- HMCRF is proposed. Most existing approaches to human motion recognition with hidden states employ a Hidden Markov Model or suitable variant to model motion streams; a significant limitation of these models is the requirement of conditional independence of observations. In contrast, conditional models like the CRFs seamlessly represent contextual dependencies, support efficient, exact inference using dynamic programming, and their parameters can be trained using convex optimization. We introduce conditional graphical models as complementary tools for human motion recognition and present an extensive set of experiments that show the proposed approach to outperform the linear-chain structure CRF and Hidden Markov Models (HMM) in terms of recognition rates.