Abstract

Persistent photoconductors of npnp doping-modulated amorphous silicon multilayers are investigated for the implementation of variable analog and nonvolatile synaptic weights of an optoelectronic neural processor. The time dependence of the persistent photoconductance of the amorphous silicon multilayers is characterized in experiments. The learning performance of an optical pattern classifier with the persistent photoconductive array is analyzed by computer simulations.

© 1991 Optical Society of America

Amorphous silicon photoconductive arrays for artificial neural networks

Edward A. Rietman, R. C. Frye, C. C. Wong, and C. D. Kornfeld
Appl. Opt. 28(16) 3474-3478 (1989)

Dynamic optical neurochip using variable-sensitivity photodiodes

J. Ohta, Y. Nitta, and K. Kyuma
Opt. Lett. 16(10) 744-746 (1991)

Hybrid optoelectronic neural networks using a mutually pumped phase-conjugate mirror

G. J. Dunning, Y. Owechko, and B. H. Soffer
Opt. Lett. 16(12) 928-930 (1991)

Optoelectronic analogs of self-programming neural nets: architecture and methodologies for implementing fast stochastic learning by simulated annealing

Nabil H. Farhat
Appl. Opt. 26(23) 5093-5103 (1987)

Dual-scale topology optoelectronic processor

Gary C. Marsden, Ashok V. Krishnamoorthy, Sadik C. Esener, and Sing H. Lee
Opt. Lett. 16(24) 1970-1972 (1991)