Combining Classifier for Face Identification at Unknown Views with a Single Model Image

We investigate a number of approaches to pose invariant face recognition. Basically, the methods involve three sequential functions for capturing nonlinear manifolds of face view changes: representation, view-transformation and discrimination. We compare a design in which the three stages are optimized separately, with two techniques which establish the overall transformation by a single stage optimization process. In addition we also develop an approach exploiting a generic 3D face model. A look-up table of facial feature correspondence between different views is applied to an input image, yielding a virtual view face. We show experimentally that the four methods developed individually outperform the classical method of Principal Component Analysis(PCA)-Linear Discriminant Analysis(LDA). Further performance gains are achieved by combining the outputs of these face recognition methods using different fusion strategies.

[1]  David J. Kriegman,et al.  From Few to Many: Illumination Cone Models for Face Recognition under Variable Lighting and Pose , 2001, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  David Beymer,et al.  Face recognition from one example view , 1995, Proceedings of IEEE International Conference on Computer Vision.

[3]  Hyeong-Seok Ko,et al.  Performance-driven muscle-based facial animation , 2001, Comput. Animat. Virtual Worlds.

[4]  Shaogang Gong,et al.  Constructing facial identity surfaces in a nonlinear discriminating space , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[5]  Jiri Matas,et al.  On Combining Classifiers , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Ian Craw,et al.  Synthetic Images of Faces — An Approach to Model-Based Face Recognition , 1991 .

[7]  J Kittler,et al.  Discriminant analysis by multiple locally linear transformations , 2003 .

[8]  G. Baudat,et al.  Generalized Discriminant Analysis Using a Kernel Approach , 2000, Neural Computation.

[9]  Jiri Matas,et al.  XM2VTSDB: The Extended M2VTS Database , 1999 .

[10]  Christoph von der Malsburg,et al.  Analysis and synthesis of human faces with pose variations by a parametric piecewise linear subspace method , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[11]  Alex Pentland,et al.  View-based and modular eigenspaces for face recognition , 1994, 1994 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[12]  Ralph Gross,et al.  Eigen light-fields and face recognition across pose , 2002, Proceedings of Fifth IEEE International Conference on Automatic Face Gesture Recognition.

[13]  Nigel M. Allinson,et al.  Automatic Face Representation and Classification , 1998, BMVC.

[14]  Sami Romdhani,et al.  Face identification across different poses and illuminations with a 3D morphable model , 2002, Proceedings of Fifth IEEE International Conference on Automatic Face Gesture Recognition.

[15]  David Casasent,et al.  Pose-invariant recognition of faces at unknown aspect views , 1999, IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339).

[16]  Terence Sim,et al.  The CMU Pose, Illumination, and Expression (PIE) database , 2002, Proceedings of Fifth IEEE International Conference on Automatic Face Gesture Recognition.

[17]  Tomaso A. Poggio,et al.  Linear Object Classes and Image Synthesis From a Single Example Image , 1997, IEEE Trans. Pattern Anal. Mach. Intell..

[18]  Hyun-Chul Kim,et al.  Discriminant Analysis by Locally Linear Transformations , 2003, BMVC.