论文信息 - Some learning models of visual system based on local sensory signals integration learning

Some learning models of visual system based on local sensory signals integration learning

The visual system has many interesting abilities like vestibule-ocular reflex. These abilities are helpful for us to perceive the environment effectively and often have adaptability to the environment. LSSI (Local Sensory Signals Integration) learning based on the smoothness of the relation between the spatial and temporal information, has been already proposed. In this paper, some computational learning models, each of which is extension of the LSSI learning, are proposed. Some of the abilities of our visual system, those of head invariant perception, vestibule-ocular reflex and smooth pursuit eye movements, are explained by those models. Basic functions are realized in some simulations using those models.

Yoichi Okabe | K. Shibata

[1] R. Leigh,et al. The neurology of eye movements , 1984 .

[2] Geoffrey E. Hinton,et al. Learning representations by back-propagating errors , 1986, Nature.

[3] Geoffrey E. Hinton,et al. Learning representations by back-propagation errors, nature , 1986 .

[4] Katsunari Shibata,et al. Unsupervised learning method to extract object locations from local visual signals , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).