Share Email Print
cover

Proceedings Paper

Optoelectronic neurocomputer based on digital optical vector-matrix multiplier
Author(s): Nickolay N. Evtikhiev; Vladimir N. Lazarev; Boris N. Onyky; Vadim V. Perepelitsa; Igor B. Scherbakov
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

The proposed optoelectronic neurocomputer consists of digital optical vector-matrix multiplier (OVMM) and general purpose electronic processor (GPEP). The basis of OVMM is a multichannel acousto-optic modulator (MAOM). OVMM performs integer vector-matrix multiplication using modified DMAC algorithm. It corresponds to data transmission through synaptic interconnections of neural net (NN). GPEP calculates values of transmission functions of neurons and prepares data for entering in OVMM. The proposed element at the base of the OVMM optoelectronic neurocomputer permits us to realize NNs of arbitrary dimension, integer values of NN parameters, with completely interconnected and multilayer architectures of interconnections. Such NNs can have important application and can be used in image recognition systems, systems of adaptive steering, etc.

Paper Details

Date Published: 21 January 1994
PDF: 6 pages
Proc. SPIE 2051, International Conference on Optical Information Processing, (21 January 1994); doi: 10.1117/12.166063
Show Author Affiliations
Nickolay N. Evtikhiev, Moscow Engineering Physics Institute (Russia)
Vladimir N. Lazarev, Moscow Engineering Physics Institute (Russia)
Boris N. Onyky, Moscow Engineering Physics Institute (Russia)
Vadim V. Perepelitsa, Moscow Engineering Physics Institute (Russia)
Igor B. Scherbakov, Moscow Engineering Physics Institute (Russia)


Published in SPIE Proceedings Vol. 2051:
International Conference on Optical Information Processing
Yuri V. Gulyaev; Dennis R. Pape, Editor(s)

© SPIE. Terms of Use
Back to Top