Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Optoelectronic implementation of neural networks

¹BT Laboratories, Martlesham Health, Ipswich, United Kingdom.

International Journal of Neural Systems

|December 1, 1993

Summary

Optical connections overcome electronic neural network limitations by using spatial optics and holographic components. Experimental optoelectronic networks demonstrate high-speed operation and novel training algorithms for future integration.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Enhanced Secondary-Electron Detection of Single-Ion Implants in Silicon through Thin SiO<sub>2</sub> Layers.

Nano letters·2026

Same author

Cosmic dust impacts on the Hubble Space Telescope.

Philosophical transactions. Series A, Mathematical, physical, and engineering sciences·2024

Same author

Effects of mating system in Japanese quail : 1. Inbreeding and fitness.

TAG. Theoretical and applied genetics. Theoretische und angewandte Genetik·2013

Same author

The fabrication of aspherical microlenses using focused ion-beam techniques.

Micron (Oxford, England : 1993)·2013

Same author

Enhanced imaging of developed fingerprints using mass spectrometry imaging.

The Analyst·2013

Same author

Integrated Ion Beam Analysis (IBA) in Gunshot Residue (GSR) characterisation.

Forensic science international·2013

Area of Science:

Optoelectronics
Artificial Neural Networks
Holographic Optics

Background:

Electronic neural network performance is limited by interconnection capacity at high speeds.
Optical connections offer a solution to overcome these limitations.
Spatial optical techniques are suitable for parallel neural network structures.

Purpose of the Study:

To describe experimental optoelectronic networks for enhanced neural network performance.
To demonstrate the feasibility of optical connections in neural systems.
To explore novel training algorithms and integration techniques.

Main Methods:

Utilizing holographic components and high-speed optical modulators.
Developing three experimental optoelectronic networks.
Implementing a network with 64 x 8 connections and optical fan-out.

Main Results:

Demonstrated network operation at 50 MHz.
Successfully employed novel training algorithms.
Developed an expansion technique for integrating optoelectronic systems.

Conclusions:

Optoelectronic networks using spatial optics can overcome performance limitations of electronic neural networks.
Holographic components and optical modulators are key to high-speed, high-capacity interconnections.
Further development in integration techniques will enable scalable optoelectronic neural systems.

Related Experiment Videos