A simple spiking retina model for exact video stimulus representation

Abstract

A computational model for the representation of visualstimuli with a population of spiking neurons is presented.We show that under mild conditions it is possible to faith-fully represent an analog video stream into a sequence ofspike trains and provide an algorithm that recovers thevideo input by using only the spike times of the popula-tion.In our model an analog, bandlimited in time, videostream approaches the dendritic trees of a neural popula-tion. At each neuron, the multi-dimensional video inputis filtered by the neuron's spatiotemporal receptive field,and the one-dimensional output dendritic current entersthe soma of the neuron (see Figure 1). The set of the spa-tial receptive fields is modeled as a Gabor filterbank. The