Facial Expression Anlysis using Eye Gaze Information

Psychologists found that human eye gaze direction has a very strong influence on facial expression. However, there exist few researches that took eye gaze direction into consideration while recognizing facial expressions. This paper makes use of the combination of eye gaze and facial expression information to detect human emotion. First, eye gaze direction is categorized into direct gaze and avert gaze. We then carry on facial expression recognition based on the eye gaze analysis results. This could provide the confusion between some of the facial expressions therefore improve the recognition rate of emotion detection. Experimental results show that our proposed method can provide more accurate recognition rate then recognizing facial expression alone.