Advertisement

Modified Predictive Validation Test for Gaussian Mixture Modelling

  • Mohammad Sadeghi
  • Josef Kittler
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2396)

Abstract

This paper is concerned with the problem of probability density function estimation using mixture modelling. In [7] and [3], we proposed the Predictive Validation, PV, technique as a reliable tool for the Gaussian mixture model architecture selection. We propose a modified form of the PV method to eliminate underlying problems of the validation test for a large number of test points or very complex models.

Keywords

Predictive Validation Gaussian Mixture Modelling Test Point Validation Test Gaussian Component 
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.

References

  1. 1.
    H. Akaike. A new look at the statistical model identification. IEEE trans. on Automatic Control, AC-19(6):716–723, 1974.CrossRefMathSciNetGoogle Scholar
  2. 2.
    A. Dempster, N. Laird, and D. Rubin. Maximum likelihood from incomplete data via the em algorithm. Journal of the Royal Statistical Society, 39(1):1–38, 1977.zbMATHMathSciNetGoogle Scholar
  3. 3.
    J. Kittler, K. Messer, and M. Sadeghi. Model validation for model selection. In S. Singh, N. Murshed, and W. Kropatsch, editors, Proceedings of International Conference on Advances in Pattern Recognition ICAPR 2001, pages 240–249, 11–14 March 2001.Google Scholar
  4. 4.
    K. Messer, J. Matas, J. Kittler, J. Luettin, and G. Maitre. Xm2vtsdb: The extended m2vts database. In Second International Conference on Audio and Video-based Biometric Person Authentication, March 1999.Google Scholar
  5. 5.
    W. Press, B. Flanney, S. Teukolsky, and W. Vetterling. Numerical Recipes in C: The Art of Scientific Computing. Cambridge University Press, Cambridge, 2nd edition, 1992.Google Scholar
  6. 6.
    M. Sadeghi, J. Kittler, and K. Messer. Segmentation of lip pixels for lip tracker initialisation. In Proceedings IEEE International Conference on Image Processing, ICIP2001, volume I, pages 50–53, 7–10 October 2001.Google Scholar
  7. 7.
    L. Sardo and J. Kittler. Model complexity validation for pdf estimation using gaus-sian mixtures. In S. V. A.K. Jain and B. Lovell, editors, International Conference on Pattern Recognition, pages 195–197, 1998.Google Scholar
  8. 8.
    G. Schwarz. Estimating the dimension of a model. The Annals of Statistics, 6(2):461–464, 1978.zbMATHMathSciNetCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Mohammad Sadeghi
    • 1
  • Josef Kittler
    • 1
  1. 1.Centre for Vision, Speech and Signal Processing School of Electronics, Computing and MathematicsUniversity of SurreyGuildfordUK

Personalised recommendations