A Sign Language Teaching System Using Sign Language Recognition and Generation Methods

Abstract

We have developed a sign language teaching system that uses sign language recognition and generation methods to overcome three problems with current learning materials: a lack of information about non-manual gestures (facial expressions, glances, head movements, etc.), display of gestures from only one or two points of view, and a lack of feedback about the correctness of the learner’s gestures. Experimental evaluation by 24 non-hearing-impaired people demonstrated that the system is effective for learning sign language.