Event Location: NSH 1305
Bio: Laszlo Attila Jeni is a research scientist at the Carnegie Mellon University and the University of Pittsburgh. The main body of his work concentrates on machine learning for affective behavior analysis and its applications, including 3D deformable face registration, facial expression analysis and human-machine interaction. At present most of his research focuses on robust constrained local models for facial expression analysis. He received his PhD from the Electrical Engineering department of the University of Tokyo and his M.Sc. from the Eotvos Lorand University, Hungary.

Abstract: In the last decade many approaches have been proposed for automatic facial expression recognition. In this talk I will be introducing the breakthrough we are experiencing in this field. This is mainly due to two factors: first is the high quality annotated databases that have been made available to everybody, and the second is the advance of learning algorithms, most notably the advance of constrained local models (CLM).
3D shape-only information based facial expression analysis represents a relatively less studied area of research until present. The main trend is to use appearance information and use shape as a side information for the classification. Line drawings, however, can express facial expressions very well, so shape information could also be a good descriptor of emotions. 3D shape as opposed to texture is attractive for facial expression recognition since it should be robust against rotations and may be robust against light conditions.
The final scope of this talk is to show that 3D shape-only information based facial expression analysis is useful for real-life scenarios, and it can offer a viable solution as an integral part of facial expression classification algorithm.