CVPR 2011 workshop on gesture recognition

Learning Structured Models for Recognizing Human Actions
Greg Mori
Simon Fraser University, Canada

The development of automatic methods for recognizing human actions is a challenging computer vision problem.  Robust solutions to this problem would facilitate a variety of applications from image retrieval to improving safety in assisted living facilities.  In this talk I will present work towards solving this problem via the learning of structured models.  I will describe a model that uses a hidden Conditional Random Field (hCRF) to learn a representation for motion parts in conjunction with whole-body templates.  Second, a variant of this model is used for treating body pose as a latent variable for action recognition.  Finally, I will describe a model relating the actions of individuals and groups of people to provide context for action recognition.  A latent variable framework that adapts the latent connections between actions of individuals will be presented.

This is joint work with Tian Lan, Yang Wang, and Weilong Yang.