Part of Advances in Neural Information Processing Systems 4 (NIPS 1991)
Jim Keeler, David Rumelhart
We present a neural network algorithm that simultaneously performs seg(cid:173) mentation and recognition of input patterns that self-organizes to detect input pattern locations and pattern boundaries. We demonstrate this neu(cid:173) ral network architecture on character recognition using the NIST database and report on results herein. The resulting system simultaneously seg(cid:173) ments and recognizes touching or overlapping characters, broken charac(cid:173) ters, and noisy images with high accuracy.