Structural Learning from Iconic Representations

  • Herman M. Gomes
  • Robert B. Fisher
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1952)


This paper addresses the important problem of how to learn geometric relationships from sets of iconic (2-D) models obtained from a sequence of images. It assumes a vision system that operates by foveating at interesting regions in a scene, extracting a number of raw primal sketch-like image descriptions, and matching new regions to previously seen ones. A solution to the structure learning problem is presented in terms of a graph-based representation and algorithm. Vertices represent instances of an image neighbourhood found in the scenes. An edge represents a relationship between two neighbourhoods. Intra and inter model relationships are inferred by means of the cliques found in the graph, which leads to rigid geometric models inferred from the image evidence.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    I. Biederman. Human image understanding: recent research and a theory. In A. Rosenfeld, editor, Hum. and Mach. Vision II, pp 13–57. Acad. Press, 1986.Google Scholar
  2. 2.
    R. B. Fisher. From Surfaces to Objects. John Wiley and Sons, 1989.Google Scholar
  3. 3.
    R. B. Fisher and A. MacKirdy. Integrating iconic and structured matching. In Proc. of Europ. Conf. on Comp. Vision, vol. II, pp 687–698, Freiburg, June 1998.Google Scholar
  4. 4.
    H. M. Gomes. Model learning in iconic vision. PhD Thesis, Division of Informatics, Edinburgh University, to be submitted, August 2000.Google Scholar
  5. 5.
    H. M. Gomes, R. B. Fisher, and J. Hallam. A retina-like image representation of primal sketch features extracted using a neural network approach. In Proc. of Noblesse Workshop on Non-Linear Model Based Image Analysis, pp 251–256, Glasgow, July 1998.Google Scholar
  6. 6.
    T. D. Grove and R. B. Fisher. Attention in iconic object matching. In Proc. of Brit. Machine Vision Conf., vol. 1, pp 293–302, Edinburgh, 1996.Google Scholar
  7. 7.
    D. Marr. Vision. W. H. Freeman and Co., 1982.Google Scholar
  8. 8.
    A. R. Pearce, T. Caelli, and W. Bischof. Rulegraphs for graph matching in pattern recognition. Patt. Recog., 27(9): 1231–1247, 1994.CrossRefGoogle Scholar
  9. 9.
    A. Pentland, R. W. Picard, and S. Sclaroff. Photobook-content-based manipulation of image databases. Int. Journal of Comp. Vision, 18(3): 233–254, 1996.CrossRefGoogle Scholar
  10. 10.
    R. P. N. Rao and D. H. Ballard. An active vision architecture based on iconic representations. Artif. Intell., 78: 461–505, 1995.CrossRefGoogle Scholar
  11. 11.
    G. Sandini and M. Tristarelli. Vision and space-variant sensing. In H. Wechsler, editor, Neural Netw. for Percep., vol. 1, chap. II.11, pp 398–425. Acad. Press, 1992.Google Scholar
  12. 12.
    B. Schiele and J. L. Crowley. Probabilistic object recognition using multidimensional receptive field histogram. In Proc. Int. Conf. on Patt. Recog., vol. B, pp 50–54, Vienna, August 1996.Google Scholar
  13. 13.
    G. Vosselman. Relational Matching. Number 628 in Lecture Notes in Computer Science. Springer-Verlag, Berlin Heidelberg, 1992.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2000

Authors and Affiliations

  • Herman M. Gomes
    • 1
  • Robert B. Fisher
    • 1
  1. 1.Division of Informatics, Edinburgh UniversityEdinburghScotland UK

Personalised recommendations