| |
Abstract:
A neural network that models the mammalian early visual
system is introduced to gain insight into possible mechanisms
operating during binocular vision. In the model, the desynchronized
firing of cortical neurons that first combine inputs from the two
eyes produces rivalrous activity patterns at later stages in the
visual pathway. By contrast, synchronization of firing among these
cells prevents such competition. The temporal coordination of
cortical activity and its effects on neural competition emerge
naturally from the network connectivity and from its dynamics.
These results suggest that input-related differences in relative
spike timing at an early stage of visual processing may give rise
to the phenomena both of perceptual integration and rivalry in
binocular vision.
|
