Multilayer recurrent neural networks [character recognition application example]

In this paper, multilayer recurrent neural networks in the form of 3-layer bidirectional symmetrical and asymmetrical associative memories are presented. The networks possess the features of both a multilayer feedforward neural network and a bidirectional associative memory. These networks can have two modes of recalling; namely, recalling by one pattern and recalling by a pattern pair. Simulation results on alpha numeral type pattern recognition performance under noisy conditions are given.