Multilayer optical learning networks
- Creators
- Wagner, Kelvin
- Psaltis, Demetri
Abstract
A new approach to learning in a multilayer optical neural network based on holographically interconnected nonlinear devices is presented. The proposed network can learn the interconnections that form a distributed representation of a desired pattern transformation operation. The interconnections are formed in an adaptive and self-aligning fashioias volume holographic gratings in photorefractive crystals. Parallel arrays of globally space-integrated inner products diffracted by the interconnecting hologram illuminate arrays of nonlinear Fabry-Perot etalons for fast thresholding of the transformed patterns. A phase conjugated reference wave interferes with a backward propagating error signal to form holographic interference patterns which are time integrated in the volume of a photorefractive crystal to modify slowly and learn the appropriate self-aligning interconnections. This multilayer system performs an approximate implementation of the backpropagation learning procedure in a massively parallel high-speed nonlinear optical network.
Additional Information
© 1987 Optical Society of America. Received 28 May 1987. The authors would like to acknowledge the numerous contributions to this work made by David Brady as well as useful discussions with Jeff Yu and Hyatt Gibbs. The work reported here was partially supported by DARPA, the Army Research Office, and the Air Force Office of Scientific Research.Attached Files
Published - WAGao87.pdf
Files
Name | Size | Download all |
---|---|---|
md5:f122c793b5c38009230230610a3e0674
|
3.1 MB | Preview Download |
Additional details
- Eprint ID
- 32080
- Resolver ID
- CaltechAUTHORS:20120626-090038005
- Defense Advanced Research Projects Agency (DARPA)
- Army Research Office (ARO)
- Air Force Office of Scientific Research (AFOSR)
- Created
-
2012-06-28Created from EPrint's datestamp field
- Updated
-
2021-11-09Created from EPrint's last_modified field