NEURAL-NETWORK APPROACHES TO VISUAL-MOTION PERCEPTION

This paper concerns certain difficult problems in image processing and perception:neurocomputation of visual motion information. The first p art of this paper deals with the spatial physiological integration by the figure-ground discrimination neural network in the visual system o f the fly. We have outlined the fundamental organization and algorithm s of this neural network, and mainly concentrated on the results of co mputer simulations of spatial physiological integration. It has been s hown that the gain control mechanism, the nonlinearity of synaptic tra nsmission characteristic, the interaction between the two eyes, and th e directional selectivity of the pool cells play decisive roles in the spatial physiological integration. In the second part, we have presen ted a self-organizing neural network for the perception of visual moti on by using a retinotopic array of Reichardt's motion detectors and Ko honen's self-organizing maps. It has been demonstrated by computer sim ulations that the network is able to learn to solve the ambiguities gi ven by local motion detection mechanism. The resultant self-organized configuration in the output layer is resembling direction selective co lumns which first appear in area MT of the primate visual system. It h as been explored that the spatio-temporal coherences, mapping, coopera tion, competition, and Hebb rule are the basic neural principles for v isual motion perception.