The Emergence of Visual Categories - A Computational Perspective

  • Pietro Perona
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2749)


When we are born we do not know about sailing boats, frogs, cell-phones and wheelbarrows. By the time we reach school age we can easily recognize these categories of objects and many more using our visual system; by some estimates, we learn around 10 new categories per day with minimal supervision during the first few years of our lives. How can this happen? I will outline a computational approach to the problem of representing the visual properties of object categories, and of learning such models without supervision from cluttered images. Both static images of objects and dynamic displays such as the ones generated by human activity are handled by the theory. Its properties will be exemplified with experiments on a variety of categories.


  1. 1.
    Leung, T., Burl, M., Perona, P.: Finding faces in cluttered scenes using random labelled graph matching. In: Proc. 5th Int. Conf. on Computer Vision, Cambridge, Mass (1995) 637–644Google Scholar
  2. 2.
    Burl, M., Leung, T., Perona, P.: Face localization via shape statistics. In: Proc. Intl. Workshop on automatic face and gesture recognition, Zurich, IEEE Computer Soc. (1995) 154–159Google Scholar
  3. 3.
    Burl, M., Weber, M., Perona, P.: A probabilistic approach to object recognition using local photometry and global geometry. In: Proc. 5th Europ. Conf. Comput. Vision, Burkhardt and Neumann (Ed.), LNCS-Series Vol. 1406–1407, Springer-Verlag. (1998)Google Scholar
  4. 4.
    M. Weber, M. W., Perona, P.: Unsupervised learning of models for recognition. (In: Proc. 6th Europ. Conf. Comp. Vis., ECCV2000, Dublin, Ireland, June 2000)Google Scholar
  5. 5.
    R. Fergus, P. P., Zisserman, A.: Object class recognition by unsupervised scale-invariant learning. (In: IEEE Comp. Society Conf. on Computer Vision and Pattern Recognition, CVPR 03, Madison, WI, June 2003)Google Scholar
  6. 6.
    Burl, M., Perona, P.: Recognition of planar object classes. In: Proc. IEEE Comput. Soc. Conf. Comput. Vision and Pattern Recogn., San Francisco (1996)Google Scholar
  7. 7.
    Leung, T., Burl, M., Perona, P.: Probabilistic affine invariance for recognition. In: Proc. IEEE Conf. Computer Vision and Pattern Recognition, Santa Barbara, CA (1998) 678–684Google Scholar
  8. 8.
    Song, Y., Goncalves, L., Bernardo, E.D., Perona, P.: Monocular perception of biological motion-detection and labeling. In: Proceedings of International Conference on Computer Vision. (1999) 805–812Google Scholar
  9. 9.
    Song, Y., Goncalves, L., Perona, P.: Monocular perception of biological motion-clutter and partial occlusion. In: Proc. ECCV. Volume 2. (2000) 719–733Google Scholar
  10. 10.
    M. Weber, W. Einhaeuser, M. W., Perona, P.: Viewpoint-invariant learning and detection of human heads. (In: Proc. 4th Int. Conf. Autom. Face and Gesture Rec., FG2000, Grenoble, France, March 2000)Google Scholar
  11. 11.
    Burl, M., Perona, P.: Using hierarchical shape models to spot keywords in cursive handwriting. (In: IEEE Comp. Society Conf. on Computer Vision and Pattern Recognition, CVPR 98, Santa Barbara, CA, June 1998)Google Scholar
  12. 12.
    M. Weber, M. W., Perona, P.: Towards automatic discovery of object categories. (In: Proc. IEEE Comp. Soc. Conf. Comp. Vis. and Pat. Rec., CVPR2000, June 2000)Google Scholar
  13. 13.
    Y. Song, L. G., Perona, P.: Unsupervised learning of human motion models. (In: Proc. of NIPS 2001) 1287–1294Google Scholar
  14. 14.
    Y. Song, L. G., Perona, P.: Unsupervised learning of human motion. IEEE Trans. Pattern Anal. Mach. Intell. 25 (2003)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Pietro Perona
    • 1
  1. 1.California Institute of TechnologyPasadena

Personalised recommendations