基于侧抑制机制的一维多速度运动检测神经网络

One Dimensional Multi-Velocity Motion Detection Neural Networks Based on Lateral Inhibitory Mechanism

运动感知是生物视觉系统的一个重要组成部分．动物对自然界的认识一大部分是通过运动感知．本文通过对同时具有方向和速率选择性的视觉皮层运动区神经元的侧抑制机制的模拟，提出了一种新的乘积抑制速度检测子（ＭｕｌｔｉｐｌｉｃａｔｉｖｅＩｎｈｉｂｉｔｏｒｙＶｅｌｏｃｉｔｙＤｅｔｅｃｔｏｒ－ＭＩＶＤ），其时空联合参数Ｋ决定了检测子的最优速度，基于ＭＩＶＤ的响应幅度差特性，我们构造了两个多速度融合神经网络（一个简单神经网络和一个主动神经网络）来检测一维运动速度．实验结果表明具有反馈的主动ＭＩＶＤ神经网络能够给出较好的检测结果．

Motion perception is one of the most important components of biological vision system. In this paper,by simulating the lateral inhibitory mechanism of the neurons in the motion area of visual cortex,we propose a new Multiplicative Inhibitory Velocity Detector(MIVD). The optimum velocity of this detector is determined by its spatial-temporal joint parameter K. Based on the respoase disarity of MIVD,we construct two multi velocity fusion Neural Networks(a simple one and an active one)to detect the velocity of one dimensional motion.The experiments show that the active MIVD Neural Networks with feedback have better detecting result.