Character recognition with neural assemblies

Abstract

In living beings, any pattern recognition task involves complex processes of concepts formation. We propose a model based on the principle of neural assemblies to develop internal representations of characters. Neural assemblies self-organize themselves by extracting the relevant information of the characters set depending on the previously stored characters. The principal consequences of this are that less storage capacity is needed, and it makes easer the clasiffication with a delta rule. Examples of this task are exposed, where it can be seen the way in which the assemblies store the characters, and how the classification of the internal representations is performed.