Stdp Training Of Hierarchical Spike Timing Model Of Visual Information Processing

We have developed a hierarchical spike timing neural network model in NEST simulator aimed to reproduce human decision in simplified simulated visual navigation task. The model consists of the following layers: retinal photoreceptors and ganglion cells (RGC); thalamic relay including lateral geniculate nucleus (LGN), thalamic reticular nucleus (TRN) and interneurons (IN); primary visual cortex (V1); middle temporal (MT) area; medial superior temporal (MST) area and lateral intraparietal cortex (LIP). All synaptic inter- and intra-layer connections of the initial model were static and structured according to the literature information. The present work extends the model with spike timing dependent plastic (STDP) synapses between MST and LIP layers. We investigated the possibility to train synaptic weights via STDP rule to mimic decisions taken by test subjects as well as to differentiate them according to their age.