3D object recognition using bidirectional modular networks

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


This paper investigates how the network scheme that consists of a set of bidirectional networks can be used to recognize gray-level images of 3D objects. The proposed scheme is based on the networks' ability to both compress and generate the input images. The paper also presents a bidirectional relaxation method that can be used to identify transformed or distorted images of 3D objects. We demonstrate through computer experiments that the proposed recognition model can learn to classify gray-level images of 3D objects and exhibit flexible alignment for transformed images.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Buhmann, J., Lange, J., von der Malsburg, C., Vorbrüggen, J.C. and Würtz, R.P. (1991). Object recognition in the dynamic link architecture: Parallel implementation of a transputer network. In B. Kosko (eds.) Neural Networks for Signal Processing, 121–159. Englewood Cliffs, NJ: Prentice Hall.Google Scholar
  2. 2.
    Cavanagh, P. (1991). What's up in top-down processing? In Gorea, A. (eds.) Representations of vision: Trends and tacit assumptions in vision research, 295–304. Cambridge University Press, Cambridge.Google Scholar
  3. 3.
    DeMers, D. and Cottrell, G. (1993). Non-linear dimensionality reduction. In Hanson, S. J., Cowan, J. D. & Giles, C. L., (eds), Advances in Neural Information Processing Systems 5. Morgan Kaufmann Publishers, San Mateo, CA. 580–587.Google Scholar
  4. 4.
    Fujita, T., Suzuki, S. and Ando, H. (1995). A network model that recognizes 3D objects: Alignment based on regularization principle. Technical Report of IEICE, NC95 (in Japanese).Google Scholar
  5. 5.
    Kohonen, T. (1988). Self-organization and associative memory. 2nd ed. Springer-Verlag.Google Scholar
  6. 6.
    Murase, H. and Nayar, S. (1995). Visual learning and recognition of 3-D objects from appearance. International Journal of Computer Vision, 14, 5–24.Google Scholar
  7. 7.
    Oja, E. (1991). Data compression, feature extraction, and autoassociation in feedforward neural networks. In Kohonen, K. et al. (eds), Artificial Neural Networks. Elsevier Science publishers B.V., North-Holland.Google Scholar
  8. 8.
    Poggio, T. and Edelman, S. (1990). A network that learns to recognize three-dimensional objects. Nature, 343, 263.PubMedGoogle Scholar
  9. 9.
    Poggio, T., Torre, V. and Koch, C. (1985). Computational vision and regularization theory. Nature, 317, 314–319.PubMedGoogle Scholar
  10. 10.
    Suzuki, S. and Ando, H. (1995). Unsupervised classification of 3D Objects from 2D views. In Tesauro, G., Touretzky, D. & Leen, T. (eds), Advances in Neural Information Processing Systems 7. The MIT Press, Cambridge, MA., 949–956.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1996

Authors and Affiliations

  • Hiroshi Ando
    • 1
  1. 1.ATR Human Information Processing Research LaboratoriesKyotoJapan

Personalised recommendations