Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

A network that learns to recognize three-dimensional objects


THE visual recognition of three-dimensional (3-D) objects on the basis of their shape poses at least two difficult problems. First, there is the problem of variable illumination, which can be addressed by working with relatively stable features such as intensity edges rather than the raw intensity images1,2. Second, there is the problem of the initially unknown pose of the object relative to the viewer. In one approach to this problem, a hypothesis is first made about the viewpoint, then the appearance of a model object from such a viewpoint is computed and compared with the actual image3–7. Such recognition schemes generally employ 3-D models of objects, but the automatic learning of 3-D models is itself a difficult problem8,9. To address this problem in computational vision, we have developed a scheme, based on the theory of approximation of multivariate functions, that learns from a small set of perspective views a function mapping any viewpoint to a standard view. A network equivalent to this scheme will thus 'recognize' the object on which it was trained from any viewpoint.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Rent or buy this article

Get just this article for as long as you need it


Prices may be subject to local taxes which are calculated during checkout


  1. Marr, D. Vision (Freeman, San Francisco, 1982).

    Google Scholar 

  2. Poggio, T., Gamble, E. B. & Little, J. J. Science 242, 436–440 (1988).

    Article  ADS  MathSciNet  CAS  Google Scholar 

  3. Fischler, M. A. & Bolles, R. C. Commun. ACM 24, 381–395 (1981).

    Article  Google Scholar 

  4. Thompson, D. W. & Munday, J. L. in Proc. IEEE Conf. Robotics and Automation 208–220 (Raleigh, North Carolina, 1987).

    Google Scholar 

  5. Huttenlocher, O. P. & Ullman, S. in Proc. 1st Int. COnf. Computer Vision 102–111 (IEEE, Washington DC, 1987).

    Google Scholar 

  6. Lowe, D. G. Perceptual Organization and Visual Recognition (Kluwer Academic Publishers, Boston, Massuchusetts, 1986).

    Google Scholar 

  7. Ullman, S. Cognition 32, 193–254 (1989).

    Article  CAS  Google Scholar 

  8. Grimson, W. E. L. & Lozano-Perez, T. IEEE Trans. Pattern Analysis Machine Intell. 9, 469–482 (1987).

    Article  CAS  Google Scholar 

  9. Fan, T. J., Medioni, G. & Nevatia, R. in Proc. 2nd Int. Conf. Computer Vision 474–481 (Florida, IEEE, Washington DC, 1988).

    Google Scholar 

  10. Tsai R. Y. & Huang, T. S. IEEE Trans. Pattern Analysis Machine Intell. 6, 13–27 (1984).

    Article  CAS  Google Scholar 

  11. Longuet-Higgins, H. C. Nature 293, 133–135 (1981).

    Article  ADS  Google Scholar 

  12. Ullman, S. The Interpretation of Visual Motion (MIT Press, Cambridge, Massachusetts, 1979).

    Google Scholar 

  13. Koenderink, J. J. & van Doorn, A. J. Biol. Cybern. 32, 211–217 (1979).

    Article  CAS  Google Scholar 

  14. Poggio, T. & Girosi, F. Artif. Intell. Lab. Memo No. 1,140 (Artificial Intelligence Laboratory, MIT, Cambridge, 1989).

    Google Scholar 

  15. Poggio, T. & Girosi, F. Science (in the press).

  16. Tikhonov, A. N. & Arsenin, V. Y. Solutions of III-posed Problems (Winston, Washington DC, 1977).

    MATH  Google Scholar 

  17. Poggio, T., Torre, V. & Koch, C. Nature 317, 314–319 (1985).

    Article  ADS  CAS  Google Scholar 

  18. Powell, M. J. D. in Algorithms for Approximation. (eds Mason, J. C. & Cox, M. G.) (Clarendon, Oxford, 1987).

    Google Scholar 

  19. Broomhead, D. S. & Lowe, D. Complex Syst. 2, 321–355 (1988).

    Google Scholar 

  20. Rumelhart, D. E., Hinton, G. E. & Williams, R. J. Nature 323, 533–536 (1986).

    Article  ADS  Google Scholar 

  21. Perrett, D. I., Mistlin, A. J. & Chitty, A. J. Trends Neurosci. 10, 358–364 (1989).

    Article  Google Scholar 

  22. Edelman, S. & Poggio, T. Optic News 15, 8–15, May 1989.

    Article  Google Scholar 

  23. Poggio, T. & Edelman, S. Artif. Intell. Lab. Memo No. 1,181 (Artificial Intelligence Laboratory, MIT, Cambridge, 1989).

    Google Scholar 

  24. Basri, R. & Ullman, S. Artif. Intell. Lab. Memo No. 1,152 (Artificial Intelligence Laboratory, MIT, Cambridge, 1989).

    Google Scholar 

  25. Rock, I. & DiVita, J. Cognitive Psychol. 19, 280–293 (1987).

    Article  CAS  Google Scholar 

  26. Edelman, S., Bülthoff, H. & Weinshall, D. Artif. Intell. Lab. Memo No. 1,138 (Artificial Intelligence Laboratory, MIT, Cambridge, 1989).

    Google Scholar 

  27. Edelman, S. & Weinshall, D. Artif. Intell. Lab. Memo No. 1,146 (Artificial Intelligence Laboratory, MIT, Cambridge, 1989).

    Google Scholar 

  28. Jenkins, W. M., Merzenich, M. M. & Ochs, M. T. Soc. Neurosci. Abstr. 10, 665 (1984).

    Google Scholar 

  29. Edelman, G. M. & Finkel, L. in Dynamical Aspects of Neocortical Function (eds Edelman, G. M., Gall, W. E. & Cowan, W. M.) 653–695 (Wiley, New York, 1984).

    Google Scholar 

  30. Gross, C. G., Rocha-Miranda, C. E. & Bender, D. B. J. Neurophys. 35, 96–111 (1972).

    Article  CAS  Google Scholar 

  31. Perrett, D. I., Rolls, E. T. & Caan, W. Expl Brain Res. 47, 329–342 (1982).

    Article  CAS  Google Scholar 

Download references

Author information

Authors and Affiliations


Rights and permissions

Reprints and Permissions

About this article

Cite this article

Poggio, T., Edelman, S. A network that learns to recognize three-dimensional objects. Nature 343, 263–266 (1990).

Download citation

  • Received:

  • Accepted:

  • Issue Date:

  • DOI:

This article is cited by


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing