École Nationale Supérieure de
Techniques Avancées

Abstract

We propose an extension to the existing large-scale neuron network simulator. SpikeSim, used in simulating plasticity. This extension comports a stimuli generation computed with Matlab and a visual front end. The visual front end enables the SpikeSim user to starts his plasticity simulation with spiking LIF neurons that mimics the retinal ganglion cell spike trains in response to visual stimuli such as moving bars, gratings, noise, or flickering screens of uniform luminance. The input current to the retinal ganglion neurons is the result of a non-separable filtering in time and space, with a center-surround shape, reproducing some important functional biological properties. The computational retinal model has been mostly inspired by the work of Wohrer (2008) and is highly configurable.