An integrated model for evaluating the amount of data required for reliable recognition

  • Michael Lindenbaum
Object Recognition
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1035)


Many recognition procedures rely on the consistency of a subset of data features with an hypothesis, as the sufficient evidence to the presence of the corresponding object. The performance of such procedures are analyzed using a probabilistic model and provide expressions for the sufficient size of such data subsets, that, if consistent, guarantee the validity of the hypotheses with arbitrarily prespecified confidence. The analysis focuses on 2D objects and on the affine transformation class, and is based, for the first time, on an integrated model, which takes into account the shape of the objects involved, the accuracy of the data collected, the clutter present in the scene, the class of the transformations involved, the accuracy of the localization, and the confidence required in the hypotheses. Most of these factors can be quantified cumulatively by one parameter, denoted “effective similarity”, which largely determines the sufficient subset size.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [BL93]
    Ben-David, S. and M. Lindenbaum, 1993, “Localization vs. Identification of Semi-Algebraic Sets”, Proceedings of the 6th ACM Conference on Computational Learning Theory, pp. 327–336.Google Scholar
  2. [B93]
    Breuel, T.M., 1993, “Higher-order Statistics in Object Recognition”, Proc. CVPR, pp. 707–708.Google Scholar
  3. [C92]
    Cass, T.A., 1992, “Polynomial Time Object Recognition in the Presence of Clutter Occlusion, and Uncertainty”, Proceedings of ECCV, pp. 834–842.Google Scholar
  4. [GTM89]
    Gottschalk P.G., J.L. Turney, and T.N. Mudge, 1989, “Efficient Recognition of Partially Visible Objects Using a Logarithmic Complexity Matching Technique”, Int. J. of Rob. Res., 8(6), pp. 110–131.Google Scholar
  5. [GH91a]
    Grimson, W.E.L., and D.P. Huttenlocher, 1991, “On the Verification of Hypothesized Matches in Model-Based Recognition”, IEEE Trans. on Pattern Analysis and Mach. Intel., PAMI-13(12), pp. 1201–1213.Google Scholar
  6. [GHJ92]
    Grimson, W.E.L., D.P. Huttenlocher, and D.W. Jacobs, 1992, “A study of affine Matching with Bounded sensor error”, Second Europ. Conf. Comp. Vision, pp. 291–306.Google Scholar
  7. [L95]
    M. Lindenbaum, 1995, “Bounds on Shape Recognition Performance”, IEEE Transactions on Pattern Analysis and Machine Intelligence PAMI-17, No.7, pp. 666–680.Google Scholar
  8. [L93]
    M. Lindenbaum, 1993, “On the Amount of Information Required for Object Recognition”, Proceedings of the 12th International Conference on Pattern Recognition, Vol. I, pp. 726–729, 1994. Also CIS report 9329, Technion, November 1993.Google Scholar
  9. [LB94a]
    M. Lindenbaum and S. Ben-David, 1994, “Applying VC-dimension Analysis to Object Recognition”, Proceedings of the 3rd European conference on Computer Vision, pp. 239–240.Google Scholar
  10. [LB94b]
    M. Lindenbaum and S. Ben-David, 1994, “Applying VC-dimension Analysis to 3D Object Recognition from Perspective Projections”, Proceedings of the 12th National Conf. on Artificial Intelligence (AAAI), pp. 985–990.Google Scholar
  11. [M93]
    Maybank, S.J., 1993, Probabilistic Analysis of the Application of the Cross Ratio to Model Based Vision”, International Journal of Computer Vision 16, pp. 5–33.Google Scholar
  12. [SG93]
    Sarachik, K.B., and W.E.L. Grimson, 1993, “Gaussian Error Models for Object Recognition”, Proc. CVPR, pp. 400–406.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1996

Authors and Affiliations

  • Michael Lindenbaum
    • 1
  1. 1.Computer Science DepartmentTechnionHaifaIsrael

Personalised recommendations