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

  • Tae-Kyun Kim
  • Josef Kittler
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3138)


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.


Face Recognition Linear Discriminant Analysis Face Image Base Classifier Face Identification 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Vetter, T., Poggio, T.: Linear object classes and image synthesis from a single example image. IEEE Trans. PAMI 19(7), 733–742 (1997)Google Scholar
  2. 2.
    Li, Y., Gong, S., Liddell, H.: Constructing facial identity surfaces in a nonlinear discriminating space. In: Proc. of CVPR (2001)Google Scholar
  3. 3.
    Graham, D.B., Allinson, N.M.: Automatic face representation and classification. In: Proc. of British Machine Vision Conference (1998)Google Scholar
  4. 4.
    Blanz, V., Romdhani, S., Vetter, T.: Face identification across different poses and illuminations with a 3D morphable model. Automatic Face and Gesture Recognition (2002)Google Scholar
  5. 5.
    Kim, T.-K., Kittler, J., Kim, H.-C., Kee, S.-C.: Discriminant analysis by multiple locally linear transformations. In: British Machine Vision Conference, Norwich, UK, pp. 123–132 (2003)Google Scholar
  6. 6.
    Talukder, A., Casasent, D.: Pose-invariant recognition of face at unknown aspect views. IEEE Joint Conf. on Neural Networks 5, 3286–3290 (1999)Google Scholar
  7. 7.
    Baudat, G., Anouar, F.: Generalized discriminant analysis using a kernel approach. Neural Computation 12, 2385–2404 (2000)CrossRefGoogle Scholar
  8. 8.
    Aitchison, A.C., Craw, I.: Synthetic images of faces – an approach to model-based face recognition. In: British Machine Vision Conference, pp. 226–232 (1991)Google Scholar
  9. 9.
    Beymer, D., Poggio, T.: Face Recognition From One Example View. In: Proc. of ICCV, pp. 500–507 (1995)Google Scholar
  10. 10.
    Kim, T.-K.: View-transformation of face images in kernel space-comparative study,Technical Report, Samsung AIT (2003)Google Scholar
  11. 11.
    Choe, B., Lee, H., Ko, H.-S.: Performance-driven muscle-based facial animation. The Journal of Visualization and Computer Animation 12, 67–79 (2001)zbMATHCrossRefGoogle Scholar
  12. 12.
    Sim, T., Baker, S., Bsat, M.: The CMU pose, illumination, and expression (PIE) database. In: Proc. of Automatic Face and Gestures Recognition (2002)Google Scholar
  13. 13.
    Okada, K., Malsburg, C.v.d.: Analysis and synthesis of human faces with pose variations by a parametric piecewise linear subspace method. In: Proc. of CVPR, vol. 1, pp. I -761–768 (2001)Google Scholar
  14. 14.
    Pentland, A., Moghaddan, B., Starner, T.: View-based and modular eigenspaces for face recognition. In: Proc. of CVPR, pp. 84–91 (1994)Google Scholar
  15. 15.
    Gross, R., Matthews, I., Baker, S.: Eigen light-fields and face recognition across pose. In: Proc. of Automatic face and Geature Recognition (2002)Google Scholar
  16. 16.
    Georghiades, A.S., Belhumeur, P.N., Kriegman, D.J.: From few to many: illumination cone models for face recognition under varialbe lighting and pose. IEEE Trans. on PAMI 23(6), 643–660 (2001)Google Scholar
  17. 17.
    Kittler, J., Hatef, M., Duin, R.P.W., Matas, J.: On combining classifiers. IEEE Trans. PAMI 20(3), 226–239 (1998)Google Scholar
  18. 18.
    Li, S., Yan, J., Hou, X., Li, Z., Zhang, H.: Learning low dimensional invariant signature of 3-D object under varying view and illumination from 2-D appearances. In: Proc. of ICCV, vol. 1, pp. 635–640 (2001)Google Scholar
  19. 19.
    Messer, K., Matas, J., Kittler, J., Lueettin, J., Maitre, G.: XM2VTSDB: The extended M2VTS database. In: Prof. of Audio and Video-based Biometric Person Authentication (1999)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2004

Authors and Affiliations

  • Tae-Kyun Kim
    • 1
  • Josef Kittler
    • 2
  1. 1.HCI LabSamsung Advanced Institute of TechnologyYonginKorea
  2. 2.Centre for Vision, Speech and Signal ProcessingUniversity of SurreyGuildfordUK

Personalised recommendations