A STUDY ON MOUTH SHAPE-FEATURES SUITABLE FOR HMM SPEECH RECOGNITION USING FUSION OF VISUAL AND AUDITORY INFORMATION

Recently, some speech recognition methods using fusion of visual and a uditory information have been researched. In this paper, a study on th e mouth shape image suitable for fusion of visual and auditory informa tion has been described. Features of mouth shape which are extracted f rom gray level image and binary image are adopted, and speech recognit ion using linear combination method has been performed. From results o f speech recognition, the studies on the mouth shape features which ar e effective in fusion of visual and auditory information have been per formed. And the effectiveness of using two kinds of mouth shape featur es also has been confirmed.