Events Calendar

PhD Defense: Analysis and Modeling of Multi-Level Dynamics of Multimodal Behavior in Affective Human Interactions

Speaker: Zhaojun Yang, University of Southern California, PhD Candidate

Talk Title: Analysis and Modeling of Multi-Level Dynamics of Multimodal Behavior in Affective Human Interactions

Abstract: Human communication is a dynamical and interactive process that is established on a common ground of conveying emotions, achieving the interaction goals and sharing mutual interests of the interaction participants. Such an interactive process naturally induces a multi-level dynamical flow along various verbal and nonverbal behavior dimensions of spoken words, speech prosody, body gestures, and facial expressions. As one of the major components that shape the structure of social interactions, emotions greatly affect the multi-level dynamics of multimodal behavior throughout the course of an interaction. This thesis, from three perspectives, explores computational methodologies to understand, analyze and model human behaviors dynamics that relate to and arise from affective processes underlying human interactions: 1) modeling the dynamics of body gesture expression of emotions; 2) studying how multimodal behavior channels, speech and body particularly, of an individual dynamically interact with one another towards emotion expression; and 3) exploring interpersonal coordination of multimodal behavior induced in human interactions.

Defense committee: Prof. Shrikanth Narayanan (Chair), Prof. C.-C. Jay Kuo, Prof. Gayla Margolin (Outside member)

Biography: Zhaojun Yang is a PhD candidate in Electrical Engineering at the University of Southern California (USC). She received her B.E. Degree from University of Science and Technology of China (USTC) 2009 and M.Phil. Degree from Chinese University of Hong Kong (CUHK) 2011. She was awarded with the USC Annenberg Fellowship (2011-2015). Her work (with S. S. Narayanan) has won the Best Student Paper Award at IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2016. Her research interests include Affective Computing, Machine Learning, and Human-centered multimodal signal processing.

Host: Shrikanth Narayanan

Location: Ronald Tutor Hall of Engineering (RTH) - 320

Audiences: Everyone Is Invited

Contact: Tanya Acevedo-Lam/EE-Systems