Human Action Recognition Using a Hidden Markov Model

Pau-Choo Chung

Department of Electrical Engineering

National Cheng-Kung University

Tainan, Taiwan

Contact email: pcchung@eembox.ncku.edu.tw

Abstract

This talk presents human actions recognition using a hidden Markov model. Human action recognition plays one of the key roles to smart human-computer interaction, one of the major goals of the next computing evolution, and human psychological behavior understanding, for studying children of psychological problem tendency. It could also be used for patient behavior monitor to prevent danger action from happening. However, human action understanding is a high-level image understanding, which remains a very challenging task. Discussed in this presentation are the applications of human action recognition and a hidden Markov model approach for identifying human actions and tracking the human’s next movement. Some experimental results of the human action recognition will also be demonstrated.