Many of our cognitive abilities, state and intentions can be revealed from our behaviors. In this talk we will present a general framework for human motion and behavior analysis that is based on facial and body movement estimations. We will first present an automated facial tracker to track facial features for expression analysis. This includes eye tracking and multiscale body motion analysis. We will then present the use of these methods in various important applications such as deception detection, emotion recognition, autism, alcohol studies, space applications and sign language. Our long term goal is to develop a theory for human behavior analysis and its relationship to brain activity.
This is collaborative work with: M. Bates (Rutgers), J. Burgoon (UA), C. Neidle (BU), E. Torres (Rutgers), D. Dinges (UPENN), B. Zucker (Michigan)
PhD Students: N. Michael, S. Zhang, M. Dilsizian, F. Yang and P. Yang