Facial expression interpretation, recognition and analy- sis is a key issue in visual communication and man to ma- chine interaction. In this paper, we address the issues of fa- cial expression recognition and synthesis and compare the proposed bilinear factorization based representations with previously investigated methods such as linear discriminant analysis and linear regression. We conclude that bilinear factorization outperforms these techniques in terms of cor- rect recognition rates and synthesis photorealism especially when the number of training samples is restrained.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below