Abstract
Higher-order networks, particularly second-order translation-invariant networks, are introduced, and their suitability for optical implementation is outlined. The algorithm is implemented with a conventional liquid-crystal display, permitting on-line learning and updating of weights. The basic operation of the optical system is demonstrated, and the ability of the system to adapt to system nonuniformities is illustrated. The implementation with an integrated optoelectronic array of asymmetric Fabry–Perot modulators containing a GaAs/AlGaAs multiple-quantum-well active region is described. The principles of operation and operating characteristics of the device array are outlined. The use of the array in an optical system to calculate the autocorrelation matrix necessary for a second-order network is demonstrated.
© 1993 Optical Society of America
Full Article | PDF ArticleMore Like This
Sunao Kakizaki, Paul Horan, Akira Arimoto, Hiroshi Sako, Atsushi Saito, and Fumio Kugiya
Appl. Opt. 33(35) 8270-8280 (1994)
Andrew Jennings, Paul Horan, and John Hegarty
Appl. Opt. 33(8) 1469-1476 (1994)
Edwin Ghahramani and L. Roger B. Patterson
Appl. Opt. 32(35) 7225-7232 (1993)