Skip Ribbon Commands Skip to main content
Help (new window)
University of Balamand > Academics > Research > Seminars > Bouchra N. Abboud

Appearance based face analysis, synthesis and recognition

SUBMITTED BY: Bouchra Nakad Abboud
ENST Paris, France

ABSTRACT: This presentation will address the issue of face representations for facial expression recognition and synthesis. In this context, a global appearance model is used in conjunction with bilinear factorization.

A feature extraction technique inspired from the above representations is then proposed which consists in automatically computing the optimal identity and expression components that best adapt to an unknown target face. This representation allows to immediately factorize the appearance of an unknown target face into an expression factor and an identity factor and supersedes the costly AAM gradient matrix construction and iterative search.

Facial expression recognition and synthesis are finally performed using each representation and their performances are compared quantitatively and qualitatively. Factorization based models yield very interesting synthesis performances in terms of visual quality of the synthetic faces. The described method is also employed to perform talking faces cloning and recognition.
​
facebook    LinkedIn    youTube    Social Media
University of Balamand,
Balamand Al Kurah,
Lebanon

Tel:  +961-6-930250
Fax: +961-6-930278