Integrating Person Identity and Expression Information for Robust Facial Expression Recognition (特集テーマ顔・表情・ジェスチャの認識・合成--表情)

概要

論文の詳細を見る
A face deformed by its expression and the change of expression depend on individuals. We proposed an integration of face identification and facial expression recognition to eleminate the effect of individual expression. A face is modelled as graph where the nodes are places at the facial feature points. These model is used for automatic face and facial feature points detection, and facial feature points tracking by applying a flexible features matching. Face identification is performed by comparing the graphs representing the input image and individual face models. Face expression are modelled by finding the relationship between the motion of facial feature points and the change of expression using B-spline curves. Individual and average expression models are generated and will be used to recognize facial expression and the degree of expression change. The expression model use in facial expression recognition is choosen by the face identification.
社団法人電子情報通信学会の論文
1999-11-19