시변패턴의 인식을 위한 2단 구조의 신경회로망

초록

This paper introduces Two-stage neural net- work that is capable of recognizing spatio-temporal patterns. First stage takes a spatio-temporal pattern as input and compress it into sparse spatio-temporal pattern. Second stage is for temporal pattern recognition with non- uniform inhibitory connections and different cell sizes. These are basic properties for detecting an embeded pattern in a larger pattern. The net- work is evaluated by computer simulation.