Self-organization of spatio-temporal multi-rhythms in neural networks stimulated by external activities

A brain model to memorize periodical signals from external stimuli is described. A living body is always being given various stimuli by the external environment. Some of these stimuli are reconstructed by spatio-temporal pattern generators in the brain. They can be thought as one of the sources of the rhythmic behaviors, such as circadian rhythm. The proposed model is based on the multi-layer type neural network with feedback connections. It can generate arbitrary periodic patterns by training of the network using external input signal. In this study, memorizing of periodic scalar signals and periodic image signals were examined. As results of computer simulations, applicability and capability of this model were confirmed.