Regular article
Parallel implementation and capabilities of entropy-driven artificial neural networks

https://doi.org/10.1016/0743-7315(92)90069-YGet rights and content

Abstract

It is shown how the recently introduced Entropy-Driven Artificial Neural Network Model (EDANN) can be implemented on a parallel transputer array, using a simulation environment which makes all decomposition and mapping issues transparent. Then, using the parallel simulator, the EDANN's capabilities are exemplified in the case of orientation extraction from retinal images. By means of simulations on the parallel machine, it is shown that the EDANN is able to adapt itself optimally to the stimulus it receives, and that the same network topology is able to accomplish both 1-D and 2-D orientation inference tasks.

References (40)

  • G. Fox
  • S. Geman et al.

    Stochastic relaxation, the Gibbs distribution and the Bayesian restoration of images

    IEEE Trans. PAMI

    (1984)
  • J. Gosh et al.

    Mapping neural networks onto message-passing multiprocessors

    J. Parallel Distrib. Comput.

    (1989)
  • S. Grossberg

    Cortical dynamics of three-dimensional form, color and brightness perception. I. Monocular theory

    Percept. Psychophys.

    (1987)
  • J.J. Hopfield

    Neural networks and physical systems with emergent collective computation abilities

  • J.J. Hopfield et al.

    Neural computations of decisions in optimization problems

    Biol. Cybernet.

    (1985)
  • INMOS Ltd.
  • B. Kamgar-Parii et al.

    On problem solving with neural networks

    Biol. Cybernet.

    (1990)
  • M. Kass

    Snakes: Active contour models

  • C. Koch et al.

    The synaptic veto mechanism: Does it underlie direction and orientation selectivity in the visual cortex?

  • Cited by (0)

    View full text