Advertisement

Analysis-by-Synthesis by Learning to Invert Generative Black Boxes

  • Vinod Nair
  • Josh Susskind
  • Geoffrey E. Hinton
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5163)

Abstract

For learning meaningful representations of data, a rich source of prior knowledge may come in the form of a generative black box, e.g. a graphics program that generates realistic facial images. We consider the problem of learning the inverse of a given generative model from data. The problem is non-trivial because it is difficult to create labelled training cases by hand, and the generative mapping is a black box in the sense that there is no analytic expression for its gradient. We describe a way of training a feedforward neural network that starts with just one labelled training example and uses the generative black box to “breed” more training data. As learning proceeds, the training set evolves and the labels that the network assigns to unlabelled training data converge to their correct values. We demonstrate our approach by learning to invert a generative model of eyes and an active appearance model of faces.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Lee, Y., Terzopoulos, D., Waters, K.: Realistic modeling for facial animation. In: SIGGRAPH (1995)Google Scholar
  2. 2.
    Sifakis, E., Neverov, I., Fedkiw, R.: Automatic determination of facial muscle activations from sparse motion capture marker data. In: SIGGRAPH (2005)Google Scholar
  3. 3.
    Moriyama, T., Kanade, T., Xiao, J., Cohn, J.F.: Meticulously detailed eye region model and its application to analysis of facial images. IEEE Transactions on Pattern Analysis and Machine Intelligence 28(5), 738–752 (2006)CrossRefGoogle Scholar
  4. 4.
    Cootes, T.F., Edwards, G.J., Taylor, C.J.: Active appearance models. IEEE Transactions on Pattern Analysis and Machine Intelligence 23(6), 681–685 (2001)CrossRefGoogle Scholar
  5. 5.
    Jordan, M., Rumelhart, D.: Forward models: Supervised learning with a distal teacher. Cognitive Science 16, 307–354 (1992)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Vinod Nair
    • 1
  • Josh Susskind
    • 1
  • Geoffrey E. Hinton
    • 1
  1. 1.Department of Computer ScienceUniversity of TorontoTorontoCanada

Personalised recommendations