Facial Event Mining Using Coupled Hidden Markov Model (TA-P2)

Author(s) : Limin Ma (Ohio University, USA) Qiang Zhou (Ohio University, USA) Mehmet Celenk (Ohio University, USA) David Chelberg (Ohio University, USA)

Abstract : Facial event mining is one of the key techniques for automatic human face analysis. It plays an important role in human computer interaction. This paper proposes a new approach to facial event recognition by means of active shape models (ASMs) and coupled hidden Markov models (CHMMs). Based on the assumption that a complex facial event can be decomposed into multiple coupled processes, ASMs are used for tracking global facial features and decoupling pattern features for upper and lower faces separately. These two interacting processes are modeled as a CHMM for training and recognition. Four basic facial events are investigated. Preliminary experiments give consistent results and show the significant advantage of CHMM over conventional hidden Markov models for facial event mining in video.

Menu