Advertisement

Image registration with shape mixtures

  • Simon Moss
  • Edwin R. Hancock
Poster Session C: Compression, Hardware & Software, Image Databases, Neural Networks, Object Recognition & Reconstruction
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1311)

Abstract

This paper describes how mixtures of Gussins be used for multiple shape template registration . The EM algorithm is applied to the shape mixture model to compute both maximum likelihood registration parameters together with set of a posteriori matching probabilities. This architecture can be viewed as providing simultaneous registration and hypothesis verification. The different templates compete to account for data through the imposed probability normalisation. Based on a sensitivity study, our main conclusions are the method is both robust to added noise and poor initialisation.

Keywords

Match Probability Posteriori Probability Maximisation Step Ambiguous Figure Expectation Step 
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.
    Bialek W. and M. Deweese, “Random switching and optimal processing in in the perception of ambiguous figures-a neural network model”, Physical Review Letters, 74, pp. 3077–3080, 1995.CrossRefPubMedGoogle Scholar
  2. 2.
    Callari F.G. and F.P. Ferrie, “Active recognition: using uncertainty to reduce ambiguity ”, Proceedings IEEE CVPR Conference, pp. 701–707, 1996Google Scholar
  3. 3.
    Dempster A.P., Rubin N.M. and Rubin D.B., “Maximum-likelihood from incomplete data via the EM algorithm”, J. Royal Statistical Soc. Ser. B (methodological),39, pp 1–38, 1977.Google Scholar
  4. 4.
    Feldman J.A. and D.H. Ballard, “Connectionist models and their Properties”, Cognitive Science, 6, pp. 205–254, 1982.CrossRefGoogle Scholar
  5. 5.
    Jordan M.I. and Jacobs R.A. “Hierarchical Mixtures of Experts and the EM Algorithm”, Neural Computation, 6, pp. 181–214, 1994.Google Scholar
  6. 6.
    Kawabata N. and T. Mori, “Disambiguating ambiguous figures by a model of selective attention”, Biological Cybernetics, 67, pp. 417–425, 1992.CrossRefPubMedGoogle Scholar
  7. 7.
    Minka T. and Picard R., “Registering Interactive Learning with a Society of Models”, IEEE Computer Society Computer Vision and Pattern Recognition Conference, pp. 447–452, 1996.Google Scholar
  8. 8.
    Moss S. and E.R. Hancock, “Registering Incomplete Radar Images using the EM Algorithm”, Proceedings of the British Machine Vision Conference, pp. 685–694, 1996.Google Scholar
  9. 9.
    Moss S. and E.R. Hancock, “Cartographic Matching onto Millimetre Radar Images”, IEEE Computer Society Workshop on Applications of Computer Vision, 1996.Google Scholar
  10. 10.
    Qin C. and J.Y.S. Luh, “Ambiguity Reduction in Relaxation Labelling”, Pattern Recognition, 27, pp. 165–180, 1994.CrossRefGoogle Scholar
  11. 11.
    Riani M. and E. Simonotto, “Stochastic resonance in the perceptual interpretation of ambiguous figures — a neural network approach”, Physical Review Letters, 72, pp. 3120–3123, 1994.CrossRefPubMedGoogle Scholar
  12. 12.
    Yuille A.L., Stolorz P. and Utans J., “Statistical Physics, Mixtures of Distributions, and the EM Algorithm” Neural Computation, 6, pp. 334–340, 1994.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • Simon Moss
    • 1
  • Edwin R. Hancock
    • 1
  1. 1.Department of Computer ScienceUniversity of YorkYorkUK

Personalised recommendations