Orientation selectivity tuning of a spike timing neural network model of the first layer of the human visual cortex

Abstract

The paper deals with the influence of some parameters determining spatial structure of a spike timing neural network model of the first layer of the human visual cortex on its orientation selectivity. For this aim the model was implemented in NEST simulator and a recently proposed approach for spatial structure design of the orientation columns in its recurrent layer was adopted. The aim was to tune the model to recognize spacial orientation of moving through the visual field stimuli with different size and orientation. The values of the parameters defining columns position and thickness as well as the photo-receptors size and variance of the LGN neurons receptive fields were determined in dependence on the stimuli characteristics. The obtained results showed that bigger size stimuli were detected by wider receptive fields while orientation of smaller stimuli was properly recognized by thicker orientation columns.