An optoelectronic neural network

Mark A.A. Neil, Ian H. White, John E. Carroll

Research output: Contribution to journalArticle

Abstract

We describe and present results of an optoelectronic neural network processing system. The system uses an algorithm based on the Hebbian learning rule to memorise a set of associated vector pairs. Recall occurs by the processing of the input vector with these stored associations in an incoherent optical vector multiplier using optical polarisation rotating liquid crystal spatial light modulators to store the vectors and an optical polarisation shadow casting technique to perform multiplications. Results are detected on a photodiode arrayarray and thresholded electronically by a controlling microcomputer. The processor is shown to work in autoassociative and heteroassociative modes with up to 10 stored memory vectors of length 64 (equivalent to 64 neurons) and a cycle time of 50ms. We discuss the limiting factors at work in this system, how they affectaffect its scalability and the general applicability of its principles to other systems.

Original languageEnglish
Pages (from-to)409-417
Number of pages9
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume1151
DOIs
Publication statusPublished - 5 Feb 1990

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Cite this

An optoelectronic neural network. / Neil, Mark A.A.; White, Ian H.; Carroll, John E.

In: Proceedings of SPIE - The International Society for Optical Engineering, Vol. 1151, 05.02.1990, p. 409-417.

Research output: Contribution to journalArticle

@article{fce7b94be9444c768d0ed679bec30060,
title = "An optoelectronic neural network",
abstract = "We describe and present results of an optoelectronic neural network processing system. The system uses an algorithm based on the Hebbian learning rule to memorise a set of associated vector pairs. Recall occurs by the processing of the input vector with these stored associations in an incoherent optical vector multiplier using optical polarisation rotating liquid crystal spatial light modulators to store the vectors and an optical polarisation shadow casting technique to perform multiplications. Results are detected on a photodiode arrayarray and thresholded electronically by a controlling microcomputer. The processor is shown to work in autoassociative and heteroassociative modes with up to 10 stored memory vectors of length 64 (equivalent to 64 neurons) and a cycle time of 50ms. We discuss the limiting factors at work in this system, how they affectaffect its scalability and the general applicability of its principles to other systems.",
author = "Neil, {Mark A.A.} and White, {Ian H.} and Carroll, {John E.}",
year = "1990",
month = "2",
day = "5",
doi = "10.1117/12.962241",
language = "English",
volume = "1151",
pages = "409--417",
journal = "Proceedings of SPIE - The International Society for Optical Engineering",
issn = "0277-786X",
publisher = "SPIE",

}

TY - JOUR

T1 - An optoelectronic neural network

AU - Neil, Mark A.A.

AU - White, Ian H.

AU - Carroll, John E.

PY - 1990/2/5

Y1 - 1990/2/5

N2 - We describe and present results of an optoelectronic neural network processing system. The system uses an algorithm based on the Hebbian learning rule to memorise a set of associated vector pairs. Recall occurs by the processing of the input vector with these stored associations in an incoherent optical vector multiplier using optical polarisation rotating liquid crystal spatial light modulators to store the vectors and an optical polarisation shadow casting technique to perform multiplications. Results are detected on a photodiode arrayarray and thresholded electronically by a controlling microcomputer. The processor is shown to work in autoassociative and heteroassociative modes with up to 10 stored memory vectors of length 64 (equivalent to 64 neurons) and a cycle time of 50ms. We discuss the limiting factors at work in this system, how they affectaffect its scalability and the general applicability of its principles to other systems.

AB - We describe and present results of an optoelectronic neural network processing system. The system uses an algorithm based on the Hebbian learning rule to memorise a set of associated vector pairs. Recall occurs by the processing of the input vector with these stored associations in an incoherent optical vector multiplier using optical polarisation rotating liquid crystal spatial light modulators to store the vectors and an optical polarisation shadow casting technique to perform multiplications. Results are detected on a photodiode arrayarray and thresholded electronically by a controlling microcomputer. The processor is shown to work in autoassociative and heteroassociative modes with up to 10 stored memory vectors of length 64 (equivalent to 64 neurons) and a cycle time of 50ms. We discuss the limiting factors at work in this system, how they affectaffect its scalability and the general applicability of its principles to other systems.

UR - http://www.scopus.com/inward/record.url?scp=84957507038&partnerID=8YFLogxK

U2 - 10.1117/12.962241

DO - 10.1117/12.962241

M3 - Article

VL - 1151

SP - 409

EP - 417

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

SN - 0277-786X

ER -