Associative dynamics of color images in a large-scale chaotic neural network

In this paper, we report a way to store color images in a large-scale chaotic neural network and to retrieve them by using chaotic dynamics. In the proposed method, color images are converted to binary codes, modified slightly by inverting a few bits, and stored in the network. The results of numerical simulations show that chaotic transitions among stored patterns and their reverse patterns can be observed within a certain range of parameters. We also compare five different coding schemes of color information, which change the appearance of chaotic dynamics. In addition, if connections are restricted in a neighborhood of each unit, a variety of wave patterns are observed.