ECCV 2004 Tutorial T4

Date: Monday May 10, 2004 afternoon Venue: Zofin Palace

Face Recognition and Modeling

Instructors: Wen-Yi Zhao, Thomas Vetter, Sami Romdhani

Duration: 4 hours

Summary:

This tutorial aims to bring the vision community a critical review on the state-of-the-art algorithms and systems in face recognition and detailed descriptions of 3D face modeling.

In this tutorial, two classes of face recognition techniques will be covered: image face recognition and video face recognition. The pros and cons of each approach will be highlighted with presentation of representative works. Recent evaluation results on face recognition systems will be discussed. In addition, we review recent advances on 3D face recognition and neuroimaging study of face recognition.

The major challenge facing today's face recognition system is to identify face images at any pose and under any illumination conditions. The state of the art system able to deal with such variations is the 3D Morphable Model. This tutorial will cover two major issues of the 3D Morphable Models: its construction and an analysis-by-synthesis framework for recognition. In the first part, the construction and the concepts motivating one of the most successful 3D face models will be explained. Then, the recognition framework based on fitting the 3D Morphable Model to any photograp in an analyzing-by-synthesizing fashion will be introduced. Different fitting strategies will be explained and compared.

Topics will include:

On the instructors:
Wen-Yi Zhao is working on vision problems related to registration, video enhancements and biometrics at the Vision Technologies Lab of Sarnoff Corporation.

Thomas Vetter is a professor of applied computer science at the University of Basel in Switzerland. His current research is on image understanding, graphics, and automated model building.

Sami Romdhani is at the University of Basel working with Thomas Vetter on fitting algorithms for 3D Morphable Models and on face recognition.

References:
[1]. Face Recognition: A Literature Survey, Wen-Yi Zhao, Rama Chellappa, P.J. Jonathon Phillips, and Azriel Rosenfeld, ACM Computing Survey, Vol. 35, No 4, 399-458, 2003.
[2]. A Morphable Model for the Synthesis of 3D Faces. Blanz, V. and Vetter, T.   SIGGRAPH'99 Conference Proceedings, pp. 187-194
[3] Face Recognition Based on Fitting a 3D Morphable Model, Blanz, V. and Vetter, T. IEEE Transactions on Pattern Analysis and Machine Intelligence,Vol. 25, No. 9, 2003.
[4] Efficient, Robust and Accurate Fitting of a 3D Morphable Model, Romdhani S. and Vetter T., Proceedings of the IEEE International Conference on Computer Vision 2003.