Bodily expression as an additional channel affect moods perception and analysis

Abstract

Human emotional assessment has been a subject of interest in human-robot interaction (HRI) system in recent years. During communication, people exchange information not only the content from the spoken words but also emotional through nonverbal messages. This paper propose bodily movement as an additional parameter affect human emotion analysis. Participant's upper-body motions were recorded during their narrative about self-experienced in both positive and negative mood. Bodily movement thus obtained was used to analyze quality of movement as an additional parameters for mood perceptual module. Hidden Markov model are used to verify the effectiveness of the bodily movement in emotion expression.