Collective dynamics of a system of adaptive velocity channels

Abstract

Our key assumption is that local, visual motion is represented by velocity channels (VCs), i.e. populations of neurons that respond optimally to edges or dots that move through their receptive field at some speed lying in a small range. This very general idea is suggested by the simple fact that the visual system is able to track moving objects within certain limits of velocity without motion blur, i.e. with a good signal to noise ratio [7]. Furthermore, it is supported by physiological studies [4, 5] as well as theoretical considerations [6].