NEURAL NETWORKS FOR SIGNAL PROCESSING - PROCEEDINGS OF THE 1992 IEEE WORKSHOP
Table of Contents
.po +0.9i
.ll -1.9i
.pl -1.0i
.fp 4 H
.sp 10
.ce 2
.ps 14
.vs 16
\f4Neural Networks for Signal Processing
Proceedings of the 1992 IEEE-SP Workshop\fR
.sp
.ps 10
.vs 12
The chapters in this book are based on presentations
given at the IEEE Signal Processing Society Workshop on
Neural Networks for Signal Processing held on August 31 - September 2, 1992
at Hotel Marienlyst, Helsingoer, Denmark.
.sp
The Technical Program Committee consisted of:
.sp
.in +0.5i
\fBRama Chellappa\fR, University of Maryland
.sp
\fBBradley Dickinson\fR, Princeton University
.sp
\fBTariq Durrani\fR, University of Strathclyde
.sp
\fBFrank Fallside\fR, Cambridge University
.sp
\fBKunihiko Fukushima\fR, Osaka University
.sp
\fBLee Giles\fR, NEC
.sp
\fBEsther Levin\fR, AT&T Bell Laboratories
.sp
\fBRichard Lippmann\fR, MIT Lincoln Laboratories
.sp
\fBJohn Makhoul\fR, BBN
.sp
\fBYasuo Matsuyama\fR, Ibaraki University
.sp
\fBJohn Moody\fR, Yale University
.sp
\fBErkki Oja\fR, Tokyo Institute of Technology
.sp
\fBWojtek Pryztula\fR, Hughes
.sp
\fBYoh'ichi Tohkura\fR, ATR Auditory & Visual
.ti +5
Perception Research Laboratories
.sp
\fBChristian Wellekens\fR, L&H Speechproducts
.bp
.in -0.5i
.sp 11
.ce 3
.ps 24
.vs 28
\f4NEURAL NETWORKS
FOR
SIGNAL PROCESSING
.sp
.ce 4
.ps 20
.vs 24
PROCEEDINGS
OF THE
1991
IEEE-SP WORKSHOP\fR
.sp
.ps 14
.vs 16
.ce
Edited by
.sp
.ce 2
\fBB. H. Juang\fR
\fIAT&T Bell Laboratories\fR
.sp
.ce 2
\fBS. Y. Kung\fI
Princeton University\fR
.sp
.ce 2
\fBCandace A. Kamm\fI
Bellcore\fR
.sp 3
.ps 10
.vs 12
.ce 4
Published under the sponsorship of the
IEEE Signal Processing Society
(in cooperation with the IEEE Neural Networks Council)
.bp
.ps 10
.vs 12
.sp 10
.ce
.ps 24
\f4Contents\fR
.ps 10
.vs 12
.sp 3
.in +0.5i
.nf
\fBPreface\fR
.sp 2
\fBPart 1: Theory and Modeling\fR
.sp 2
Note on Generalization, Weight Decay, and Architecture
Selection in Nonlinear Learning Systems
\fIJ. E. Moody\fR
.sp
Discriminative Multi-Layer Feed-Forward Networks
\fIS. Katagiri, C. H. Lee, and B. H. Juang\fR
.sp
Efficient Training Procedures for Adaptive
Kernel Classifiers
\fIS. V. Chakravarthy, J. Ghosh, L. Deuser, and S. Beck\fR
.sp
Concept Formation and Statistical Learning in
Nonhomogeneous Neural Nets
\fIR. L. Tutwiler and L. H. Sibul\fR
.sp
An Alternative Proof of Convergence for
Kung-Diamantaras APEX Algorithm
\fIH. Chen and R. Liu\fR
.sp
Neural Networks for Extracting Unsymmetric
Principal Components
\fIS. Y. Kung and K. I. Diamantaras\fR
.sp
The Outlier Process
\fID. Geiger and R. A. M. Pereira\fR
.sp
A Mapping Approach for Designing Neural Sub-Nets
\fIK. Rohani, M. S. Chen, and M. T. Manry\fR
.sp
Three-Dimensional Structured Networks for
Matrix Equation Solving
\fIL. X. Wang and J. M. Mendel\fR
.bp
.sp 10
Improving Learning Rate of Neural Tree Networks
Using Thermal Perceptrons
\fIA. Sankar and R. Mammone\fR
.sp
Adaline with Adaptive Recursive Memory
\fIB. De Vries, J. C. Principe, and P. Guedes de Oliveira\fR
.sp
Learned Representation Normalization:
Attention Focusing with Multiple Input Modules
\fIM. L. Rossen\fR
.sp
A Parallel Learning Filter System That Learns
the KL-Expansion from Examples
\fIR. Lenz and M. Osterberg\fR
.sp
Restricted Learning Algorithm and Its Application
to Neural Network Training
\fIT. Miyamura, I. Yamada, and K. Sakaniwa\fR
.sp
Multiply Descent Cost Competitive Neural Networks
with Cooperation and Categorization
\fIY. Matsuyama\fR
.sp
Nonlinear Adaptive Filtering of Systems with
Hysteresis by Quantized Mean Field Annealing
\fIR. A. Nobakht, S. H. Ardalan, and D. E. Van den Bout\fR
.sp
An Outer Product Neural Net for Extracting
Principal Components from a Time Series
\fIL. E. Russo\fR
.sp 2
\fBPart 2: Pattern Recognition\fR
.sp 2
Pattern Recognition Properties of Neural Networks
\fIJ. Makhoul\fR
.sp
Edge Detection for Optical Image Metrology
Using Unsupervised Neural Network Learning
\fIH. K. Aghajan, C. D. Schaper, and T. Kailath\fR
.bp
.sp 10
Improving Generalization Performance
in Character Recognition
\fIH. Drucker and Y. Le Cun\fR
.sp
Neural Networks for Sidescan SONAR
Automatic Target Detection
\fIM. J. LeBlanc and E. S. Manolakos\fR
.sp
An Effection Method for Visual Pattern
Recognition
\fII. N. M. Papadakis\fR
.sp
Fingerprint Recognition Using Neural Network
\fIW. F. Leung, S. H. Leung, W. H. Lau, and A. Luk\fR
.sp
A Comparison of Second-Order Neural Networks
to Transform-Based Methods for Translation-
and Orientation-Invariant Object Recognition
\fIR. Duren and B. Peikari\fR
.sp
Shape Recognition with Nearest Neighbor
Isomorphic Network
\fIH. C. Yau and M. T. Manry\fR
.sp
Dimensionality Reduction of Dynamical Patterns
Using a Neural Network
\fIS. Nakagawa, Y. Ono, and Y. Hirata\fR
.sp
A Critical Overview of Neural Network Pattern
Classifiers
\fIR. Lippmann\fR
.sp 2
\fB Part 3: Speech Processing\fR
.sp 2
Workstation-Based Phonetic Typewriter
\fIT. Kohonen\fR
.sp
Word Recognition with the Feature Finding
Neural Network (FFNN)
\fIT. Gramss\fR
.bp
.sp 10
New Discriminative Training Algorithms Based
on the Generalized Probabilistic Descent Method
\fIS. Katagiri, C. H. Lee, and B. H. Juang\fR
.sp
Probability Estimation by Feed-Forward Networks
in Continuous Speech Recognition
\fIS. Renals, N. Morgan, and H. Bourlard\fR
.sp
Nonlinear Resampling Transformation
for Automatic Speech Recognition
\fIY. D. Liu, Y. C. Lee, H. H. Chen, and G. Z. Sun\fR
.sp
Speech Recognition by Combining Pairwise
Discriminant Time-Delay Neural Networks and
Predictive LR-Parser
\fIJ. Takami, A. Kai, and S. Sagayama\fR
.sp
Speech Recognition Using Time-Warping Neural
Networks
\fIK. Aikawa\fR
.sp
A Hybrid Continuous Speech Recognition System
Using Segmental Neural Nets with Hidden Markov
Models
\fIS. Austin, G. Zavaliagkos, J. Makhoul, and R. Schwartz\fR
.sp
Connectionist Speaker Normalization and Its
Application to Speech Recognition
\fIX. D. Huang, K. F. Lee, and A. Waibel\fR
.sp
A Time-Derivative Neural Net Architecture -
An Alternative to the Time-Delay Neural Net
Architecture
\fIK. K. Paliwal\fR
.sp
Word Recognition Based on the Combination
of a Sequential Neural Network and
the GPDM Discriminative Training Algorithm
\fIW. Y. Chen and S. H. Chen\fR
.sp
A Space-Perturbance/Time-Delay Neural Network
for Speech Recognition
\fIM. Ji, H. H. Chen, and Z. K. Shen\fR
.bp
.sp 10
Non-Linear Prediction of Speech Signals Using
Memory Neuron Networks
\fIP. Poddar and K. P. Unnikrishnan\fR
.sp
Experiments with Temporal Resolution for
Continuous Speech Recognition with
Multi-Layer Perceptrons
\fIN. Morgan, C. Wooters, and H. Hermansky\fR
.sp
Neural-Network Architecture for Linear and
Nonlinear Predictive Hidden Markov Models:
Applications to Speech Recognition
\fIL. Deng, K. Hassanein, and M. Elmasry\fR
.sp
On Adaptive Acquisition of Spoken Language
\fIA. L. Gorin, S. E. Levinson, L. G. Miller & A. N. Gertner\fR
.sp
Vector Quantisation with a Codebook-Excited
Neural Network
\fIL. Wu and F. Fallside\fR
.sp
Segment-Based Speaker Adaptation by Neural
Network
\fIK. Fukuzawa, H. Sawai, and M. Sugiyama\fR
.sp
A Simple Word-Recognition Network with the
Ability to Choose Its Own Decision Criteria
\fIK. A. Fischer and H. W. Strube\fR
.sp
Supervised and Unsupervised Feature Extraction
from a Cochlear Model for Speech Recognition
\fIN. Intrator and G. Tajchman\fR
.sp 2
\fBPart 4: Signal Processing\fR
.sp 2
A Relaxation Neural Network Model for Optimal
Multi-Level Image Representation
by Local-Parallel Computations
\fIN. Sonehara\fR
.bp
.sp 10
Lithofacies Determination from Wire-Line Log
Data Using a Distributed Neural Network
\fIM. Smith, N. Carmichael, I. Reid & C. Bruce\fR
.sp
Improved Structures Based on Neural Networks
for Image Compression
\fIS. Carrato, G. Ramponi, A. Premoli, and G. L. Sicuranza\fR
.sp
Adaptive Neural Filters
\fIL. Yin, J. Astola, and Y. Neuvo\fR
.sp
A Surface Reconstruction Neural Network
for Absolute Orientation Problems
\fIJ. N. Hwang and H. Li\fR
.sp
Recursive Neural Networks for Signal Processing
and Control
\fID. Hush, C. Abdallah, and B. Horne\fR
.sp
A Neural Architecture for Nonlinear Adaptive
Filtering of Time Series
\fIN. Hoffmann and J. Larsen\fR
.sp
Ordered Neural Maps and Their Applications
to Data Compression
\fIE. A. Riskin, L. E. Atlas, and S. R. Lay\fR
.sp
Vector Quantization of Images Using Neural
Networks and Simulated Annealing
\fIM. Lech and Y. Hua \fR
.sp
A Multilayer Perceptron Feature Extractor
for Reading Sequenced DNA Autoradiograms
\fIM. Murdock, N. Cotter, and R. Gesteland\fR
.sp
Configuring Stack Filters by the LMS Algorithm
\fIN. Ansari, Y. Huang, and J. H. Lin\fR
.sp
A Neural Network Pre-Processor for Multi-Tone
Detection and Estimation
\fIS. S. Rao and S. Sethuraman\fR
.bp
.sp 10
Fuzzy Tracking of Multiple Objects
\fIL. I. Perlovsky\fR
.sp 2
\fBPart 5: System Implementation\fR
.sp 2
Neural Nets for Signal/Image Processing
Using the Princeton Engine Multiprocessor
\fIN. Binenbaum, L. Dias, P. Hsieh, J. Ju, S. Markel,
J. C. Pearson, and H. Taylor, Jr.\fR
.sp
Design of a Digital VLSI Neuroprocessor for Signal
and Image Processing
\fIC. F. Chang and B. Sheu\fR
.sp
Tutorial: Digital Neurocomputing for Signal/Image
Processing
\fIS. Y. Kung\fR
.sp 2
\fBAuthor Index\fR
.bp
.in -0.5i
.sp 11
.fi
.ce
.ps 24
\f4Preface\fR
.sp 2
.ps 10
.vs 12
This book contains papers presented at the IEEE Workshop on Neural
Networks for Signal Processing (NNSP-91) at Princeton, New Jersey,
USA on September 30 - October 2, 1991. This is the first
workshop on the subject sponsored by the IEEE Signal
Processing Society, in cooperation with the IEEE Neural Networks
Council.
.sp
The workshop, organized by the
Neural Network Technical Committee of the IEEE Signal Processing Society, is
designed to serve as a regular forum for researchers from
universities and industry who are interested in interdisciplinary
research on neural networks for signal
processing applications.
In the present scope, the workshop encompasses
up-to-date results in several key areas, including
learning theory, neural models, speech processing,
signal processing, image processing,
pattern recognition, and system implementation.
This Conference
Proceedings is crafted to be an archival reference in the rapidly
growing field of Neural Networks for Signal Processing.
.sp
Our deep appreciation is extended to Professor Teuvo Kohonen, Helsinki
University of Technology, Helsinki, Finland, for his keynote address
titled "Workstation-Based Phonetic Typewriter", and to Dr. J.
Makhoul, BBN Systems & Technologies, Cambridge, MA., USA, for his
keynote address titled "Pattern Recognition Properties of Neural
Networks".
Our sincere thanks go to all the authors for their timely contributions and
to all the members of the Program Committee for the outstanding and
high-quality program.
Also, we would like to express our gratitude to
Dr. John Vlontzos for taking care of the local
arrangements, to Dr. Gary Kuhn for providing the workshop publicity,
to Dr. Bastiaan Kleijn for handling the tedious finance and
registration matters, and to all the session chairs for their help
in making the workshop a success.
Finally, we are indebted to Ms. Susan Gafgen and Ms. Kim Hegelbach of
Princeton University for their invaluable assistance in organizing
the workshop.
.sp
.in +3.0i
.nf
\fIB. H. Juang
.sp
S. Y. Kung
.sp
Candace A. Kamm\fR
.fi
.bp
.in -3.0i
.sp 18
.ps 24
.vs 26
.ce 4
\f4Part 1:
.sp
Theory
and
Modeling\fR
.bp
.ps 10
.vs 12
.sp 18
.ps 24
.vs 26
.ce 3
\f4Part 2:
.sp
Pattern
Recognition\fR
.bp
.ps 10
.vs 12
.sp 18
.ps 24
.vs 26
.ce 3
\f4Part 3:
.sp
Speech
Processing\fR
.bp
.ps 10
.vs 12
.sp 18
.ps 24
.vs 26
.ce 3
\f4Part 4:
.sp
Signal
Processing\fR
.bp
.ps 10
.vs 12
.sp 18
.ps 24
.vs 26
.ce 3
\f4Part 5:
.sp
System
Implementation\fR