EigenSegments: A Spatio-Temporal Decomposition of an Ensemble of Images

  • Shai Avidan
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2352)


Eigensegments combine image segmentation and Principal Component Analysis (PCA) to obtain a spatio-temporal decomposition of an ensemble of images. The image plane is spatially decomposed into temporally correlated regions. Each region is independently decomposed temporally using PCA. Thus, each image is modeled by several low-dimensional segment-spaces, instead of a single high-dimensional image-space. Experiments show the proposed method gives better classification results, gives smaller reconstruction errors, can handle local changes in appearance and is faster to compute. Results for faces and vehicles are shown.


Principal Component Analysis Support Vector Machine Test Image Receiver Operator Characteristic Reconstruction Error 
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.
    C. M. Bishop and J. M. Winn. Non-linear Bayesian Image Modelling. In European Conference on Computer Vision. Dublin 2000.Google Scholar
  2. 2.
    C. Bregler and S. M. Omohundro. Nonlinear manifold learning for visual speech recognition. In Fifth International Conference on Computer Vision, pages 494–499, Boston, June 1995.Google Scholar
  3. 3.
    R. O. Duda and P. E. Hart. Pattern Classification and Scene Analysis. Wiley-Interscience publication, 1973.Google Scholar
  4. 4.
    B. J. Frey, A. Colmenarez and T. S. Huang. Mixtures of Local Linear Subspaces for Face Recognition. In IEEE Conference on Computer Vision and Pattern Recognition, Sant Barabara, CA, June 1998.Google Scholar
  5. 5.
    G. E. Hinton, M. Revow and P. Dayan. Recognizing handwritten digits using mixtures of linear models. In Advances in Neural Information Processing Systems 7, G. Tesauro, D. Touretzky and T. Leen, Eds. 19955, pp. 1015–1022, MIT Press.Google Scholar
  6. 6.
    B. Moghaddam and A. Pentland. Probabilitic visual learning for object recognition. In IEEE Transactions on Pattern Ananlysis and Machine Inteliligence, 19(7):696–710, 1997.CrossRefGoogle Scholar
  7. 7.
    Y. Moses. The Weizmann facebase,
  8. 8.
    A. Pentland, B. Moghaddam and T. Starner. View-based and Modular Eigenspaces for Face Recognition. In em IEEE Conference on Computer Vision and Pattern Recognition, Seattle, WA, June, 1994.Google Scholar
  9. 9.
    F. Samaria and A. Harter. Parameterisation of a stochastic model for human face identification. In 2nd IEEE Workshop on Applications of Computer Vision December 1994, Sarasota (Florida).Google Scholar
  10. 10.
    L. Sirovich and M. Kirby. Low-dimensional procedure for the characterization of human faces. In Journal of the Optical Society of America 4, 510–524.Google Scholar
  11. 11.
    M. Turk and A. Pentland. Eigenfaces for recognition. In Journal of Cognitive Neuroscience vol. 3, no. 1, 1991.Google Scholar
  12. 12.
    V. Vapnik. The Nature of Statistical Learning Theory. Springer, N.Y., 1995.zbMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Shai Avidan
    • 1
  1. 1.Kanfei NesharimThe Interdisciplinary CenterHerzliyaIsrael

Personalised recommendations