Analysis of Facial Expressions and Attractiveness Using Facial Feature Information

Abstract

This paper presents a case study of facial expression and attractiveness analysis that uses facial feature information. In Experiment I, the FACS judgments revealed that the persistence of emotion and facial expression affected the straight face and smile. In Experiment II, the difference between morning and night was clarified by calculating the needle shape degree of the eye fissure, which is a geometric feature of the face. It was found that the upper eyelid muscle was affected by the level of arousal rather than muscle fatigue. In Experiment III, the location of moles was investigated and the relationship between their location and impressions was analyzed. Impression evaluation and principal component analysis resulted in an image scale with the first principal component being attractiveness and the second principal component being steadiness. By using different facial feature information, we can use it as a guideline for make-up and impression management.