Neural network simulation of the early visual system

A retinal model is presented to simulate the response of the early visual system to changes in intensity and position using neuronal networks. The model is capable of identifying a stimulus on a screen and specifying if any changes occur from one time period (tau) to the next. The neural structures used in the model are the photoreceptor cells, amacrine cells, and the X, Y, and W ganglion cells. The input is made of a N*N matrix. This retinal model faithfully replicates the receptive field characteristics of the vertebrate early visual system, as well as the general functions of the ganglion cells.<>