Abstract

Electronic neural networks can perform the function of associative memory. Given an input pattern, the network searches through its stored memories to find which of them best matches the input. Thus the network does a combination of content-addressable search and error correction. The number of random memories that a network can store is limited to a fraction of the number of electronic neurons in the circuit. We propose a method for building a hierarchy of networks that allows the fast parallel search through a list of memories that is too large to store in a single network. We have demonstrated the principle of this approach by an example in image vector quantization.

© 1987 Optical Society of America

Electronic neural network chips

L. D. Jackel, H. P. Graf, and R. E. Howard
Appl. Opt. 26(23) 5077-5080 (1987)

Dynamic control of an artificial neural system: the property inheritance network

Thomas W. Ryan, C. L. Winter, and Charles J. Turner
Appl. Opt. 26(23) 4961-4971 (1987)

Training of a neural network for image superresolution based on a nonlinear interpolative vector quantizer

Carlos A. Dávila and B. R. Hunt
Appl. Opt. 39(20) 3473-3485 (2000)

Electronic hardware implementations of neural networks

A. P. Thakoor, A. Moopenn, John Lambe, and S. K. Khanna
Appl. Opt. 26(23) 5085-5092 (1987)

Optical associative memory model based on neural networks having variable interconnection weights

Bohdan Macukow and Henri H. Arsenault
Appl. Opt. 26(5) 924-928 (1987)