A neural network model of programming for the saccade goal

The human fovea is a very small spot. In order to look at wide visual fields, we need saccadic eye movements. Nevertheless saccadic eye movements exist, humans can perceive exact and motion targets. Therefore eye movement is deeply concerning with visual recognition. So we propose a neural network model of programming saccade by which we may be able to explain such behavior of visual perception and recognition. Robinson et al. (1975) proposed a model of the oculomotor system. In this paper we discuss previous models and confirm that saccadic eye movements are globally controlled with the feedforward system. We try to apply of it to explain visual information processing mechanisms.