Browsing by Author "HSU, KY"

Now showing 1 - 13 of 13

A HYBRID NEURAL-NETWORK FOR IMAGE CLASSIFICATION
(1995)
HSU, KY
;
LIN, SH
;
電子物理學系
;
光電工程學系
;
Department of Electrophysics
Principles of the photorefractive perceptron learning algorithm are described. The influences of the finite response time and hologram erasure of the photorefractive gratings on the convergence property of the photorefractive perceptron learning are discussed. A novel neural network which could resolve these constraints is presented. It is a hybrid system which utilizes the photorefractive holographic gratings to implement the inner product between the input image and the interconnection matrix. A personal computer is used for storing the interconnection matrix and the updating procedure, and it also functions as a feedback means during the learning phase. After training the weight vectors are recorded in the volume hologram of an optical processor. This novel method combines the advantages of the massive parallelism of optical systems and the programmability of electronic computers. Experimental results of image classification are presented. It shows that the system could correctly classify the input patterns into one of the two groups after training on four examples in each group during successive iterations. The system has been extended to perform multi-category image classification.
ANISOTROPIC STRONG VOLUME HOLOGRAM IN BATIO3
(1995-09-15)
SUN, CC
;
CHANG, MW
;
HSU, KY
;
光電工程學系
;
Department of Photonics
A photorefractive strong volume hologram by using anisotropic diffraction in a normal-cut BaTiO3 is proposed and demonstrated. Experimental measurements fit the theoretical calculations well by using Kogelnik's formula with unclamped E-O coefficients. The maximum refractive index perturbation in the experiment is larger than 2.7 X 10(-4). In addition, we have observed that the output of the diffracted light are concentric circles. This pattern is shown as a good guide of the estimation of how large the refractive index perturbation is. The Bragg condition of the anisotropic strong volume hologram is so strict that an image-carried hologram cannot be directly read out. We proposed an address image method by erasure for solving this problem. Through this method, this anisotropic strong volume hologram has some potential for optical image storage.
CONDITIONAL CONVERGENCE OF PHOTOREFRACTIVE PERCEPTRON LEARNING
(1993-12-15)
HSU, KY
;
LIN, SH
;
YEH, P
;
光電工程學系
;
Department of Photonics
We consider the convergence characteristics of a perceptron learning algorithm, taking into account the decay of photorefractive holograms during the process of interconnection weight changes. As a result of the hologram erasure, the convergence of the learning process is dependent on the exposure time during the weight changes. A mathematical proof of the conditional convergence, as well as computer simulations of the photorefractive perceptrons, is presented and discussed.
CONTRAST-REVERSIBLE PHOTOREFRACTIVE INCOHERENT-TO-COHERENT OPTICAL CONVERTER BY USING AN ANISOTROPIC STRONG VOLUME HOLOGRAM
(1993-04-15)
SUN, CC
;
CHANG, MW
;
HSU, KY
;
光電工程學系
;
Department of Photonics
An anisotropic strong volume hologram with a large coupling parameter is demonstrated in a normal-cut BaTiO3 crystal. We use this hologram to fabricate an incoherent-to-coherent optical converter that has a contrast-reversible output.
GENERALIZED PERCEPTRON LEARNING RULE AND ITS IMPLICATIONS FOR PHOTOREFRACTIVE NEURAL NETWORKS
(1994-09-01)
CHENG, CJ
;
YEH, PC
;
HSU, KY
;
交大名義發表
;
光電工程學系
We consider the properties of a generalized perceptron learning network, taking into account the decay or the gain of the weight vector during the training stages. A mathematical proof is given that shows the conditional convergence of the learning algorithm. The analytical result indicates that the upper bound of the training steps is dependent on the gain (or decay) factor. A sufficient condition of exposure time for convergence of a photorefractive perceptron network is derived. We also describe a modified learning algorithm that provides a solution to the problem of weight vector decay in an optical perceptron caused by hologram erasure. Both analytical and simulation results are presented and discussed.
HOLOGRAM ENHANCEMENT IN PHOTOREFRACTIVE MEDIA
(1995-08-01)
YEH, PC
;
GU, C
;
CHENG, CJ
;
HSU, KY
;
光電工程學系
We consider the possibility of restoration and/or enhancement of decaying holograms in photorefractive media by using a simple optical readout in conjunction with a phase conjugator. The results indicate that extremely weak holograms can be enhanced provided that the two-beam coupling is sufficiently strong. Steady-state photorefractive holograms can be maintained continuously without decay by using a properly designed readout scheme. The result also provides an explanation for the formation of mutually pumped phase conjugation in terms of the amplification of initial noise gratings.
MATRIX MATRIX MULTIPLICATION BY USING ANISOTROPIC SELF-DIFFRACTION IN BATIO3
(1994-07-10)
SUN, CC
;
CHANG, MW
;
HSU, KY
;
光電工程學系
;
Department of Photonics
We propose a matrix-matrix multiplication by using anisotropic self-diffraction in BaTiO3. The input matrices are carried by the two incident beams with special Bragg-matched incident angles. The output matrices are produced by anisotropic self-diffraction with the polarization orthogonal to those of the incident matrices. By thresholding the output this architecture is particularly suitable for the applications of optical interconnects and an optical switch.
MISALIGNED SPHERICAL-MIRROR WAVE-GUIDE RESONATORS
(1986-07-01)
HSIEH, TC
;
HSU, KY
;
LI, Y
;
光電工程學系
;
Department of Photonics
OPTICAL RESTORATION OF PHOTOREFRACTIVE HOLOGRAMS
(1995-11-01)
YEH, P
;
GU, C
;
CHENG, CJ
;
HSU, KY
;
光電工程學系
We consider the possibility of restoration and/or enhancement of decaying holograms in photorefractive media by using a simple optical readout in conjunction with a phase conjugator. The results indicate that extremely weak holograms can be enhanced provided that the two-beam coupling is sufficiently strong. Steady-state photorefractive holograms can be maintained continuously without decay by using a self-enhanced readout scheme. The results also provide an explanation for the formation of mutually pumped phase conjugation in terms of the amplification of an initial noise grating. The results are presented and discussed.
OPTICAL RESTORATION OF PHOTOREFRACTIVE HOLOGRAMS THROUGH SELF-ENHANCED DIFFRACTION
(1995-02-01)
CAMPBELL, S
;
YEH, PC
;
GU, C
;
LIN, SH
;
CHENG, CJ
For the first time to our knowledge, we demonstrate the use of self-enhanced diffraction to optically restore weak dynamic volume holograms. This restoration process is achieved through the use of a novel beam-toggling scheme that significantly enhances the steady-state diffraction efficiency of volume index gratings stored in photorefractive media. Experiments demonstrating and confirming theory are presented and discussed.
OPTIMAL INCIDENT CONDITIONS FOR A CAT-TYPE SELF-PUMPED PHASE-CONJUGATE MIRROR
(1992-09-20)
SUN, CC
;
YEH, S
;
CHANG, MW
;
HSU, KY
;
光電工程學系
We present a simplified method, based on experimental observations, for predicting the best incident conditions for obtaining the highest reflectivity of a Cat-type self-pumped phase conjugator with given crystal dimensions.
REAL-TIME IMAGE PROCESSING USING PHOTOREFRACTIVE CRYSTALS
(1994)
HSU, KY
;
LIN, SH
;
HSIEH, TC
;
光電工程學系
;
Department of Photonics
TRAINABLE PATTERN CLASSIFIER USING A COMPUTER-GENERATED HOLOGRAM
(1994-04-01)
HSU, KY
;
CHENG, CJ
;
HSIEH, TC
;
電子物理學系
;
光電工程學系
A novel technique for pattern classification is proposed and demonstrated. The perceptron learning algorithm is used for training, and the trained interconnection weight matrix is transferred into a computer-generated hologram. The hologram is used as a correlation filter in the optical system for pattern classification. Both computer simulation and optical experiment results are presented.