Four-quadrant optical matrix vector multiplication machine as a neural network processor

Shai Abramson, D. Saad, Emanuel Marom, Naim Konforti

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Optical processors for neural networks are primarily fast matrix-vector multiplication machines that can potentially compete with serial computers owing to their parallelism and their ability to facilitate densely connected networks. However, in most proposed systems the multiplication supports only two quadrants and is thus unable to provide bipolar neuron outputs for increasing network capabilities and learning rate. We propose and demonstrate an opto-electronic four quadrant matrix-vector multiplier that can be used for feedforward neural networks recall and learning. Experimental results obtained with common commercial components demonstrate a novel, useful, and reliable approach for four quadrant matrix-vector multiplication in general and for feedforward neural network training and recall in particular.

Original languageEnglish
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
EditorsItzhak Shladov, Yitzhak Weissman, Moshe Oron
PublisherPubl by Society of Photo-Optical Instrumentation Engineers
Pages166-174
Number of pages9
ISBN (Print)081941218X
StatePublished - 1993
Event8th Meeting on Optical Engineering on Israel: Optoelectronics and Applications in Industry and Medicine - Tel Aviv, Isr
Duration: 14 Dec 199216 Dec 1992

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume1972
ISSN (Print)0277-786X

Conference

Conference8th Meeting on Optical Engineering on Israel: Optoelectronics and Applications in Industry and Medicine
CityTel Aviv, Isr
Period14/12/9216/12/92

Fingerprint

Dive into the research topics of 'Four-quadrant optical matrix vector multiplication machine as a neural network processor'. Together they form a unique fingerprint.

Cite this