International Journal of Computer Vision

, Volume 20, Issue 3, pp 243–261 | Cite as

Exploiting the generic viewpoint assumption

  • William T. Freeman


The “generic viewpoint” assumption states that an observer is not in a special position relative to the scene. It is commonly used to disqualify scene interpretations that assume special viewpoints, following a binary decision that the viewpoint was either generic or accidental. In this paper, we apply Bayesian statistics to quantify the probability of a view, and so derive a useful tool to estimate scene parameters.

Generic variables can include viewpoint, object orientation, and lighting position. By considering the image as a (differentiable) function of these variables, we derive the probability that a set of scene parameters created a given image. This scene probability equation has three terms: the fidelity of the scene interpretation to the image data; the prior probability of the scene interpretation; and a new genericity term, which favors scenes likely to produce the observed image. The genericity term favors image interpretations for which the image is stable with respect to changes in the generic variables. It results from integration over the generic variables, using a low-noise approximation common in Bayesian statistics.

This approach may increase the scope and accuracy of scene estimates. It applies to a range of vision problems. We show shape from shading examples, where we rank shapes or reflectance functions in cases where these are otherwise unknown. The rankings agree with the perceived values.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Albert, M.K. and Hoffman, D.D. 1995. Genericity in spatial vision. In D. Luce (Ed.), Geometric Representations of Perceptual Phenomena: Papers in Honor of Tarow Indow's 70th Birthday. L. Erlbaum (in press).Google Scholar
  2. Albrecht D.G. and Geisler W.S. 1991. Motion selectivity and the contrast-response function of simple cells in the visual cortex. Visual Neuroscience, 7:531–546.Google Scholar
  3. Belhumeur, P.N. 1996. A computational theory for binocular stereopsis. In D. Knill and W. Richards (Eds.), Perception as Bayesian Inference. Cambridge University Press.Google Scholar
  4. Berger, J.O. 1985. Statistical Decision Theory and Bayesian Analysis. Springer.Google Scholar
  5. Bichsel, M. and Pentland, A.P. 1992. A simple algorithm for shape from shading. In Proc. IEEE CVPR, Champaign, IL, pp. 459–465.Google Scholar
  6. Biederman I. 1985. Human image understanding: Recent research and a theory. Comp. Vis., Graphics, Image Proc., 32:29–73.Google Scholar
  7. Binford T.O. 1981. Inferring surfaces from images. Artificial Intelligence, 17:205–244.Google Scholar
  8. Bleistein, N. and Handelsman, R.A. 1986. Asymptotic Expansions of Integrals. Dover.Google Scholar
  9. Box G.E.P. and Tiao G.C. 1964. A Bayesian approach to the importance of assumptions applied to the comparison of variances, Biometrika, 51(1, 2):153–167.Google Scholar
  10. Box, G.E.P. and Tiao, G.C. 1973. Bayesian Inference in Statlstical Analysis. John Wiley and Sons, Inc.Google Scholar
  11. Brainard, D.H. and Freeman, W.T. 1994. Bayesian method for recovering surface and illuminant properties from photosensor responses. In Proceedings of SPIE, vol. 2179, San Jose, CA.Google Scholar
  12. Brooks M.J. and Horn B.K.P. 1989. Shape and source from shading. In B.K.P.Horn and M.J.Brooks (Eds.), Shape from Shading, MIT Press: Cambridge, MA, Chap. 3.Google Scholar
  13. Brooks M.J., Chojnacki W., and Kozera R. 1992. Impossible and ambiguous shading patterns. Int. J. Comp. Vis., 7(2):119–126.MathSciNetMATHGoogle Scholar
  14. Bulthoff, H.H. 1991. Bayesian models for seeing shapes and depth. Journal of Theoretical Biology, 2(4).Google Scholar
  15. Carandini M. and Heeger D.J. 1994. Summation and division by neurons in primate visual cortex. Science, 264:1333–1336.Google Scholar
  16. Cook, R.L. and Torrance, K.E., 1981. A reflectance model for computer graphics. In SIGGRAPH-81.Google Scholar
  17. Cornsweet, T.N. 1970. Visual Perception. Academic Press.Google Scholar
  18. Darrell, T., Sclaroff, S., and A. Pentland. 1990. Segmentation by minimal description. In Proc. 3rd Intl. Conf. Computer Vision, Osaka, Japan, IEEE, pp. 112–116.Google Scholar
  19. Dickinson S.J., Pentland A.P., and Rosenfeld A., 1992. 3-d shape recovery distributed aspect matching. IEEE Pat. Anal. Mach. Intell., 14(2):174–198.Google Scholar
  20. Fisher, R.A. 1959. Statistical Methods and Scientific Inference. Hafner.Google Scholar
  21. Freeman, W.T. 1993. Exploiting the generic view assumption to estimate scene parameters. In Proc. 4th Intl. Conf. Comp. Vis., Berlin, IEEE, pp. 347–356.Google Scholar
  22. Freeman W.T. 1994. The generic viewpoint assumption in a framework for visual perception. Nature, 368(6471):542–545.Google Scholar
  23. Freeman, W.T. and Brainard, D.H. 1995. Bayesian decision theory, the maximum local mass estimate, and color constancy. In Proc. 5th Intl. Conf. Comp. Vis., Boston, IEEE, pp. 210–217.Google Scholar
  24. Freeman, W.T. 1996. The generic viewpoint assumption in a Bayesian framework. In D. Knill and W. Richards (Eds.) Perception as Bayesian Inference. Cambridge University Press.Google Scholar
  25. Geman S. and Geman D. 1984. Stochastic relaxation, Gibbs distribution, and the Bayesian restoration of images. IEEE Pat. Anal. Mach. Intell., 6:721–741.Google Scholar
  26. Grimson E. 1984. Bionocular shading and visual surface reconstruction. Comp. Vis., Graphics, Image Proc., 28:19–43.Google Scholar
  27. Gull, S.F. 1988. Bayesian inductive inference and maximum entropy. In G.J. Erickson and C.R. Smith (Eds.), Maximum Entropy and Bayesian Methods in Science and Engineering, Kluwer, vol. 1.Google Scholar
  28. Gull S.F. 1989. Developments in maximum entropy data analysis. In J.Skilling (Ed.), Maximum Entropy and Bayesian Methods, Cambridge. Kluwer, pp. 53–71.Google Scholar
  29. Heeger, D.J. and Simoncelli, E.P. 1992. Model of visual motion sensing. In L. Harris and M. Jenkin (Eds.), Spatial Vision in Humans and Robots. Cambridge University Press.Google Scholar
  30. Horn B.K.P. 1989. Height and gradient from shading, Technical Report1105, MIT Artificial Intelligence Lab. MIT, Cambridge, MA 02139.Google Scholar
  31. Horn B.K.P. and BrooksM.J. 1989. Shape from Shading. MIT Press: Cambridge, MA.Google Scholar
  32. Horn B.K.P., Szeliski R., and Yuille A. 1993. Impossible shaded images. IEEE Pat. Anal. Mach. Intell., 15(2):166–170.Google Scholar
  33. Horn B.K.P., Woodham R.J., and Silver W.M. 1978. Determining shape and reflectance using multiple images. Technical Report 490, Artificial Intelligence Lab. Memo. Massachusetts Institute of Technology, Cambridge, MA 02139.Google Scholar
  34. Human Vision, Visual Processing and Digital Display V. Google Scholar
  35. Jeffreys H. 1961. Theory of Probability. Clarendon Press: Oxford.Google Scholar
  36. Jepson, A.D. and Richards, W. 1992. What makes a good feature? In L. Harris and M. Jenkin (Eds.), Spatial Vision in Humans and Robots. Cambridge Univ. Press. See also MIT AI Memo. 1356 (1992).Google Scholar
  37. Johnson R.A. 1970. Asymptotic expansions associated with posterior distributions. The Annals of Mathematical Statistics, 41(3):851–864.Google Scholar
  38. Kersten D. 1991. Transparency and the cooperative computation of scene attributes. In M.S.Landy and J.A.Movshon (Eds.), Computational Models of Visual Processing. MIT Press: Cambridge, MA, Chapter 15.Google Scholar
  39. Knill, D.C., Kersten, D., and Yuille, A. 1996. A Bayesian formulation of visual perception. In D. Knill and W. Richards (Eds.), Perception as Bayesian Inference. Cambridge University Press.Google Scholar
  40. Koenderink J.J. and vanDoorn A.J. 1979. The internal representation of solid shape with respect to vision. Biol. Cybern., 32:211–216.Google Scholar
  41. Laplace, P.S. 1812. Theorie Analytique des Probabilites. Courcier.Google Scholar
  42. Leclerc Y.G. 1989. Constructing simple stable descriptions for image partitioning. Intl. J. Comp. Vis., 3:73–102.Google Scholar
  43. Leclerc, Y.G. and Bobick, A.F. 1991. The direct computation of height from shading. In Proc. IEEE CVPR, Maui, Hawii, pp. 552–558.Google Scholar
  44. Lee C.-H. and Rosenfeld A. 1989. Improved methods of estimating shapefrom shading using the light source coordinate system. In B.K.P.Horn and M.J.Brooks (Eds.), Shape from Shading. MIT Press: Cambridge, MA, Chapter 11.Google Scholar
  45. Lindley, D.V. 1972. Bayesian Statistics, A Review. Society for Industrial and Applied Mathematics (SIAM).Google Scholar
  46. Lowe D.G. and Binford T.O. 1985. The recovery of threedimensional structure from image curves. IEEE Pat. Anal. Mach. Intell., 7(3): 320–326.PubMedGoogle Scholar
  47. MacKay D.J.C. 1992. Bayesian interpolation. Neural Computation, 4(3): 415–447.Google Scholar
  48. Malik J. 1987. Interpreting line drawings of curved objects. Intl. J. Comp. Vis., 1:73–103.Google Scholar
  49. Nakayama K. and Shimojo S. 1992. Experiencing and perceiving visual surfaces. Science, 257:1357–1363.Google Scholar
  50. Papoulis A. 1984. Probability, Random Variables, and Stochastic Processes. McGraw-Hill: New York.Google Scholar
  51. Pentland A.P. 1984. Local shading analysis. IEEE Pat. Anal. Mach. Intell., 6(2): 170–187.Google Scholar
  52. Pentland, A.P. 1990a. Photometric motion. In Proceedings of 3rd International Conference on Computer Vision.Google Scholar
  53. Pentland A.P. 1990b. Automatic extraction of deformable part models. Intl. J. Comp. Vis., 4:107–126.Google Scholar
  54. Pentland T., Torre V. and Koch C. 1985. Computational vision and regularization theory. Nature, 317(26):114–139.Google Scholar
  55. Press, W.H., Teukolsky, S.A., Vetterling, W.T., and Flannery B.P. 1992, Numerical Recipes in C. Cambrige University Press.Google Scholar
  56. Richards W.A., Koenderink J.J., and Hoffman D.D. 1987. Inferring three-dimensional shapes from two-dimensional silhouettes. J. Opt. Soc. Am. A, 4(7):1168–1175.Google Scholar
  57. Schreiber, W.F. 1986. Fundamentals of Electronic Imaging Systems. Springer Verlag.Google Scholar
  58. Skilling J. 1989. Classic maximum entropy. In J.Skilling (Ed.), Maximum Entropy and Bayesian Methods. Cambridge, Kluwer, pp. 45–52.Google Scholar
  59. Szeliski J. 1989. Bayesian Modeling of Uncertainty in Low-Level Vision, Kluwer Academic Publishers: Boston.Google Scholar
  60. Terzopoulos D. 1986. Regularization of inverse problems involving discontinuities. IEEE Pat. Anal. Mach. Intell., 8(4):413–424.Google Scholar
  61. Tikhonov A.N. and Arsenin V.Y. 1977. Solutions of ll-Posed Problems, Winston: Washington, DC.Google Scholar
  62. Weinshall, D., Werman, M., and Tishby, N. 1994. Stability and likelihood of views of three dimensional objects. In Proceedings of the 3rd European Conference on Computer Vision, Stockholm, Sweden.Google Scholar
  63. Witkin A.P. 1981. Recovering surface shape and orientation from texture. Artificial Intelligence, 17:17–45.Google Scholar
  64. Woodham R.J. 1980. Photometric method for determing surface orientation from multiple images. Optical Engineering 19(1):139–144.Google Scholar
  65. Yuille, A.L. and Bulthoff, H.H. 1996. Bayesian decision theory and psychophysics. In D. Knill and W. Richards (Eds.), Perception as Bayesian Inference. Cambridge University Press.Google Scholar
  66. Zheng Q. and Chellapa R. 1991. Estimation of illuminant direction, albedo, and shape from shading. IEEE Pat. Anal. Mach. Intell., 13(7):680–702.Google Scholar

Copyright information

© Kluwer Academic Publishers 1996

Authors and Affiliations

  • William T. Freeman

There are no affiliations available

Personalised recommendations