Home // BIOTECHNO 2011, The Third International Conference on Bioinformatics, Biocomputational Systems and Biotechnologies // View article

Computational Modeling of Robust Figure/Ground Separation

Authors:
Marc Ebner
Stuart Hameroff

Keywords: visual perception; spiking neurons; lateral-coupling; gap-junctions; gamma-oscillations

Abstract:
It is unknown which computational method the brain uses to perceive a visual scene. Given current advancements, it is now possible to model perceptual processes of the brain using spiking neural network models. We have developed a computational model for robust figure/ground separation. The model is based on a laterally connected sheet of spiking neurons. The sheet of neurons receives its visual input from a virtual retina. It is assumed to be located inside V1 or a higher visual area. The neurons are assumed to be laterally connected to their nearest neighbors through gap-junctions. These lateral connections allow the neurons to exchange information and therefore allow for robust figure/ground separation. Even though we only show results for visual signals, the method is quite general and may be used in various areas of the brain. A result of the lateral coupling is that the neurons synchronize their firing behavior resulting in the so called gamma-synchrony which is also a result of our computational model.

Pages: 67 to 72

Copyright: Copyright (c) IARIA, 2011

Publication date: May 22, 2011

Published in: conference

ISSN: 2308-4383

ISBN: 978-1-61208-137-3

Location: Venice/Mestre, Italy

Dates: from May 22, 2011 to May 27, 2011