HANDWRITTEN ALPHANUMERIC CHARACTER-RECOGNITION BY THE NEOCOGNITRON

A neural network model of visual pattern recognition, called the neocognitron, was previously proposed by one of the present authors. It is capable of deformation-invariant visual pattern recognition. In this paper, we discuss a pattern-recognition system which works with the mechanism of the neocognitron. The system has been trained to recognize 35 hand-written alphanumeric characters. The ability to correctly recognize deformed characters depends highly on the choice of the training pattern set. This paper offers some techniques for selecting training patterns useful for deformation-invariant recognition of a large number of characters.