A Visual Perceiving and Eyeball-Motion Controlling Neural Network for Object Searching and Locating

This paper proposes a visual cognitive neural network for automatic object searching and locating. The model consists of two sub-networks. One is a visual perceiving network, which simulates human eyes to input image signals and recognize an object´s direction and distance in terms of a high-level perceiving neuron´s maximum response. The other one is an eyeball-motion controlling network, which simulates that human brain´s high-level perceiving neurons transfer their responses to eyeball-motion controlling muscle cells to change eye´s gaze to the position of the object that the perceiving system is attentive to or interested in. The system is applied to human face features searching and experiments show a promising result.