A model of computation in neocortical architecture
Neural Networks - Special issue on organisation of computation in brain-like systems
| Hi-index | 0.01 |
Spike-time-based models of visual processing often cover only a few spike waves. This disregards the continuous character of visual processing and the fact that the visual field is scanned by series of saccades to different fixation points. Thus, some form of reset is needed in order to bring the neurons into a well-defined state after each saccade. In this contribution, we investigate the ability of various tentative reset mechanisms to maintain a latency-based temporal code in a model of layer 4 spiny stellate cells in the primary visual cortex of primates. We conclude that input suppression and inhibition of neurons are effective ways of setting temporal reference frames for spike-time processing.