CaltechTHESIS
  A Caltech Library Service

Optics in neural computation

Citation

Levene, Michael (1998) Optics in neural computation. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/XQVE-SA13. https://resolver.caltech.edu/CaltechETD:etd-08132004-133148

Abstract

In all attempts to emulate the considerable powers of the brain, one is struck by both its immense size, parallelism, and complexity. While the fields of neural networks, artificial intelligence, and neuromorphic engineering have all attempted oversimplifications on the considerable complexity, all three can benefit from the inherent scalability and parallelism of optics. This thesis looks at specific aspects of three modes in which optics, and particularly volume holography, can play a part in neural computation.

First, holography serves as the basis of highly-parallel correlators, which are the foundation of optical neural networks. The huge input capability of optical neural networks make them most useful for image processing and image recognition and tracking. These tasks benefit from the shift-invariance of optical correlators. In this thesis, I analyze the capacity of correlators, and then present several techniques for controling the amount of shift invariance. Of particular interest is the Fresnel correlator, in which the hologram is displaced from the Fourier plane. In this case, the amount of shift invariance is limited not just by the thickness of the hologram, but by the distance of the hologram from the Fourier plane.

Second, volume holography can provide the huge storage capacity and high speed, parallel read-out necessary to support large artificial intelligence systems. However, previous methods for storing data in volume holograms have relied on awkward beamsteering or on as-yet non-existent cheap, wide-bandwidth, tunable laser sources. This thesis presents a new technique, shift multiplexing, which is capable of very high densities, but which has the advantage of a very simple implementation. In shift multiplexing, the reference wave consists of a focused spot a few millimeters in front of the hologram. Multiplexing is achieved by simply translating the hologram a few tens of microns or less. This thesis describes the theory for how shift multiplexing works based on an unconventional, but very intuitive, analysis of the optical far-field. A more detailed analysis based on a path-integral interpretation of the Born approximation is also derived. The capacity of shift multiplexing is compared with that of angle and wavelength multiplexing.

The last part of this thesis deals with the role of optics in neuromorphic engineering. Up until now, most neuromorphic engineering has involved one or a few VLSI circuits emulating early sensory systems. However, optical interconnects will be required in order to push towards more ambitious goals, such as the simulation of early visual cortex. I describe a preliminary approach to designing such a system, and show how shift multiplexing can be used to simultaneously store and implement the immense interconnections required by such a project.

Item Type:Thesis (Dissertation (Ph.D.))
Subject Keywords:holographic data storage; holography; neural networks; neuromorphic engineering; shift multiplexing
Degree Grantor:California Institute of Technology
Division:Engineering and Applied Science
Major Option:Computation and Neural Systems
Thesis Availability:Public (worldwide access)
Research Advisor(s):
  • Psaltis, Demetri
Thesis Committee:
  • Psaltis, Demetri (chair)
  • Tanguay, Armand
  • Fraser, Scott E.
  • Bridges, William B.
Defense Date:10 April 1998
Record Number:CaltechETD:etd-08132004-133148
Persistent URL:https://resolver.caltech.edu/CaltechETD:etd-08132004-133148
DOI:10.7907/XQVE-SA13
Default Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:3108
Collection:CaltechTHESIS
Deposited By: Imported from ETD-db
Deposited On:16 Aug 2004
Last Modified:21 Dec 2019 01:53

Thesis Files

[img]
Preview
PDF (Levene_m_1998.pdf) - Final Version
See Usage Policy.

8MB

Repository Staff Only: item control page