Abstract
We propose a new method for the optical implementation of the Hopfield neural network with a universal tool. The tool is a matrix grating constituted with a group of element gratings. The algorithms for designing a matrix grating are proposed, and a matrix grating is created to execute recognition experiments by use of the Hopfield neural network. The experimental results demonstrate that the proposed method performs well. The stability of the light efficiencies of different optical components used in optical networks is also considered.
© 2004 Optical Society of America
Full Article | PDF ArticleMore Like This
Fumihiko Ito and Ken-ichi Kitayama
Appl. Opt. 28(19) 4176-4181 (1989)
Nabil H. Farhat, Demetri Psaltis, Aluizio Prata, and Eung Paek
Appl. Opt. 24(10) 1469-1475 (1985)
H. J. White and W. A. Wright
Appl. Opt. 27(2) 331-338 (1988)