ESCA Workshop on Audio-Visual Speech Processing (AVSP'97)

September 26-27, 1997
Rhodes, Greece

Analytical Method for Linguistic Information of Facial Gestures in Natural Dialogue Languages

A. Ichikawa, Y. Okada, A. Imiya, K. Horiuchi

Department of Information and Computer Sciences, Chiba University, Japan

A new dynamic method for the analysis of facial gestures used in natural dialogue languages in order to obtain linguistic information is proposed. Sign language is a typical natural dialogue language. In sign language, it is well known that facial gestures communicate very important linguistic information (Kimura, H. 1996, etc). We try to develop a methodology for the dynamic analysis of facial gestures used during sign language. The proposed method is based on the dynamic programming pattern matching method without any distinguishing marks on the subject's face. The experimental results show the feasibility of this method for the analysis of facial gestures which contain linguistic information. The computational load of the proposed image processing method is very light.

Full Paper

Bibliographic reference.  Ichikawa, A. / Okada, Y. / Imiya, A. / Horiuchi, K. (1997): "Analytical method for linguistic information of facial gestures in natural dialogue languages", In AVSP-1997, 105-108.