Speech Processing, Recognition and Artificial Neural Networks: Proceedings of the 3rd International School on Neural Nets “Eduardo R. Caianiello”

347 Pages·1999·1.21 MB·English

by  John J. Ohala (auth.)

Speech Processing, Recognition and Artificial
Neural Networks
Springer
London
Berlin
Heidelberg
New York
Barcelona
Hong Kong
Milan
Paris
Santa Clara
Singapore
Tokyo
Gerard Chollet, Maria Gabriella Di Benedetto,
Anna Esposito and Maria Marinaro (Eds)
Speech Processing,
Recognition and
Artificial Neural
Networks
Proceedings of the 3rd International School on Neural Nets
"Eduardo R. Caianiello"
Springer
Gerard Chollet, PhD
ENST -CNR URA 820, 46 rue Barrault, 75634 Paris Cedex 13, France
Maria Gabriella Di Benedetto, PhD
1NFOCOM Department, Rome University "La Sapienza", via Eudossiana 18,100184
Rome, Italy
Anna Esposito, PhD
IIASS, via G Pellegrino 19,1-84019 Vietri sul Mare (SA), Italy
Maria Marinaro, PhD
IIASS, via G Pellegrino 19,1-84019 Vietri suI Mare (SA), Italy
TSBN-13: 978-1-85233-094-1 Springer-Verlag London Berlin Heidelberg
British Library Cataloguing in Publication Data
Speech processing, recognition and artificial neural
networks: proceedings of the 3rd International School on
Neural Nets "Eduardo R. Caianiello"
l.Speech processing systems -Congresses 2.Speech
perception -Congresses 3.N eural networks (Computer
science) -Congresses
LChollet, Gerard ILInternational School on Neural Nets
(3rd: 1998: Salerno, Italy)
006.4'54
ISBN-13: 978-1-85233-094-1 e-ISBN-13: 978-1-4471-0845-0
DOl: 10.1007/978-1-4471-0845-0
Library of Congress Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress
Apart from any fair dealing for the purposes of research or private study, or criticism or review, as
permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced,
stored or transmitted, in any form or by any means, with the prior permission in writing of the
publishers, or in the case of reprographic reproduction in accordance with the terms of licences issued by
the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent
to the publishers.
© Springer-Verlag London Limited 1999
The use of registered names, trademarks etc. in this publication does not imply, even in the absence of a
specific statement, that such names are exempt from the relevant laws and regulations and therefore free
for general use.
The publisher makes no representation, express or implied, with regard to the accuracy of the
information contained in this book and cannot accept any legal responsibility or liability for any errors or
omissions that may be made.
Typesetting: Camera ready by contributors
34/3830-543210
Organizing -Scientific Committee:
Gerard Chollet (ENST -CNRS URA, Paris, France);
Maria Gabriella Di Benedetto (INFO COM -University "La Sapienza",
Rome, Italy);
Anna Esposito (IIASS -E.R. Caianiello, Vietri sul Mare, Italy);
Maria Marinaro (Dept. of Theoretical Physics, Salerno University, and
IIASS -E.R. Caianiello, Vietri sul Mare, Italy).
The sponsorship and support of:
* International Institute for Advanced Scientific Studies "E.R. Caianiello"
(IIASS)
* European Speech Communication Association (ESCA)
* University of Salerno
are gratefully acknowledged
List of Participants
Acernese, Fausto -Universita di Salerno, Italy
Al-ghoneim, Khaled -King Saud University, Riyadh, Saudi Arabia
AI-Sulaiman, Mansour Mohammed -King Saud University, Riyadh, Saudi
Arabia
Albesano, Dario -CSELT, Torino, Italy
Alwan, Abeer -Dept. of Electrical Engineering, 66-147E IV, UCLA, Los
Angeles, USA
Amato, Daniela -IIASS, Vietri sul Mare, Italy
Auckenthaler, Roland -University of Wales, UK
Aversano, Guido -Universita di Salerno, Italy
Budillon, Alessandra -Universita di Napoli "Federico II", Naples, Italy
Calabrese, Andrea -Linguistics Dept., Storss, USA
Cassese, Maria -IIASS, Vietri sul Mare, Italy
Cellesi, Francesco -INFN, Naples, Italy
Cheboub, Leila -TCTS LAB, Faculte Polytechnique de Mons, Belgium
Chen, Fang -IKP/IAV Linkoping University, Sweden
Chollet, Gerard -CNRS URA-820, ENST, Dept. SIG, Paris, France
Cosi, Piero -Istituto di Fonetica e Dialettologia, CNR, Padova, Italy
D' Allievo, Matgherita Leondina -Universita di Pescara, Italy
D'Eugenio, Mirko -Universit' dell'Aquila, Italy
De Giovanni, Maria Teresa -IIASS, Vietri sul Mare, Italy
De Mori, Renato -Laboratoire Informatique (LIA), Univ. d'Avignon,
France
Di Benedetto, Maria G. -INFO-COM, University of Rome, Italy
Di Fabio, Mirko -Universit' dell' Aquila, Italy
Di Perna, Francesco -Universita dell' Aquila, Italy
Di Toma, Elisabetta -Universita dell' Aquila, Italy
Donato, Daniel-Universita di Salerno, Italy
Donato, I. Yuri -Universita di Salerno, Italy
Dos Santos, Camargo Laurizete -Instituto Tecnologico de Aeronautica
San Jose dos Campos, Brazil
Esposito, Anna -IIASS, Vietri sul Mare, Italy
Farinas, Jerome -Institut de Recherche en Informatique de Toulouse,
France
Giaquinta, Maria -IIASS, Vietri sul Mare, Italy
Gill Favela, Barbara -Scuola Normale Superiore, Pisa, Italy
Giovannardi, Maurizio -INFO-COM, University of Rome, Italy
Goronzy, Silke -Sony International (Europe) GmbH Stuttgart Tech.
Center, Germany
Granstrom, Bjorn -Dept. of Speech, Music and Hearing, Centre for
Speech Technology, Stockholm, Sweden
Haton, Jean-Paul- Univ. Henri Poincare, Nancy 1, Membre de Institut
Universaire de France, LOIRA/INRIA BP 239
Heldner, Mattias -Umea University, Department of Phonetics, Sweden
Hermansky, Hynek -Oregon Graduate Institute, USA
Houard, Sarah -TCTS LAB, Faculte Polytechnique de Mons, Belguim .
Izzo, Graziano -IIASS, Vietri sul Mare, Italy
Jesus, Luis -ISIS Research Group, Dept. of Electronics and Computer
Science, Southampton, UK
Lienard, Jean S. -LIMSI-CNRS, Orsay, France
Marinaro, Maria -IIASS, Vietri sul Mare and Universita di Salerno, Italy
Mattera, Davide -Universita di Napoli "Federico II", Italy
Mauuary, Laurent -Centre National d'Etudes des Telecommunications
DIHIDIPS Technopole Anticipa, Lannion, France
Ney, Hermann -Lehrstuhl flir lnformatik VI, RWTH Aachen, University
of Technology, Germany
Ohala, John J. -Dept. of Linguistics, University of California, Berkeley,
USA
Padovani, Marco -IIASS, Vietri sul Mare, Italy
Palma, Giuseppe -IIASS, Vietri sul Mare, Italy
Palmieri, Francesco -Universita di Napoli "Federico II", Italy
Riccio, Giovanni -IIASS, Vietri sul Mare, Italy
Rigotti, Camilla -Politecnico di Milano, Dip. di Bioingegneria, Italy
Rojc, Matej -University of Maribor, Maribor, Slovenia
Savino, Michelina -D.E.E Politecnico di Bari, Italy
Sazhok, Mykola -NAS Institute of Cybernetics and UNESCO/IIP, Kyjiv,
Ukraine
Tounsi, Nawfal-TCTS LAB, Faculte Polytechnique de Mons, Belguim
Vicinanza, Sergio -IIASS, Vietri sul Mare, Italy
Vignoli, Fabio -DIST, University of Genova, Italy
Vintsiuk, Taras -NAS Institute of Cybernetics and UNESCO/lIP, Kyjiv,
Ukraine
Vitulano, Domenico -lAC, Rome, Italy
Voglino, Giuseppe -ST Microelectronics Central R&D, DAIS
Weber, Andrea C. -Max-Planck-Institute for Psycho-linguistics, Holland
Preface
This volume contains invited and contributed papers presented at the
International Summer School Neural Nets "E.R. Caianiello" on Speech
Processing, Recognition, and Artificial Neural Networks", held in Vietri
sul Mare, Salerno, Italy, October 5-141998.
The aim of this book is to provide primarily high level tutorial coverage of
the fields related to speech science and linguistics, including automatic
speech recognition, connectionist approaches to speech, and Hidden
Markov Models (HMM).
Twelve surveys are offered by specialists in the field. Consequently, the
volume may be used as a reference book on speech. Six original papers,
which present original contributions and complete the tutorials are
included.
The volume is divided into five sections: Fundamentals of Speech
Analysis and Perception, Speech Processing, Stochastic Models for
Speech, Auditory and Neural Network Models for Speech, and Task
Oriented Applications of Automatic Speech Recognition.
Fundamentals of Speech Analysis and Perception deals with problems related
to the acoustic-phonetic theory in which basic speech sounds are
characterized according to both their linguistic properties and the
associated acoustic measurements. Fundamental and innovative ideas in
speech perception are covered. This section contains four tutorial papers.
The first tutorial, authored by John Ohala, discusses underlying
aerodynamic forces essential to speech sounds. The second, by Andrea
Calabrese, describes and accounts for various phonetic and phonological
processes which characterize normal connected speech. The tutorial
presented by Maria-Gabriella Di Benedetto and Anna Esposito deals with
the problem of identifying a set of acoustic and perceptual features for
characterizing basic speech sounds (vowels and consonants). Finally, the
tutorial by Jean-Sylvain Lienard reports on the physiological,
psychological and linguistic mechanisms which underlie the process of
speech perception.
Speech Processing deals with new and standard techniques used to provide
speech features used in recognition systems. This section contains two
tutorials. The first by Hynek Hermansky, reports on early speech analysis
techniques and on recent auditory modeling for speech analysis. The
second, by Abeer Alwan et al., deals with two-dimensional preprocessing
techniques which combine both articulatory and acoustic information
embedded in the speech signal.
Stochastic Models for Speech includes the tutorials by Hermann Ney (Search
Strategies and Statistical Translation) and Renato De Mori (Language
Models), and describes speech recognition systems based on HMM and
the methods used for selecting and modeling speech units. This section
also contains four contributed papers which complete the tutorials.
Auditory and Neural Network Models for Speech includes the tutorials by
Piero Cosi (Auditory Models) and Jean-Paul Haton (Neural Networks),
and provides an overview of recent applications for speech analysis and
recognition using biological models. Two contributed papers dealing
with phoneme classification using neural networks are included.
Task-Oriented Applications of Automatic Speech Recognition and Synthesis
includes the papers by Gerard Chollet et al. (Speech technology for
Telephone Interactive Voice Servers) and Bjorn Granstrom (Multi-Modal
Speech Synthesis). This section analyzes the principles by which task
oriented applications may be successful.
The editors would like to thank the European Speech Communication
Association (ESCA) and in particular professor W olfang Hess, the
International Institute for Advanced Scientific Studies "E.R. Caianiello",
and the University of Salerno for their support in sponsoring, financing,
and organizing the school.
In addition, the editors are grateful to the contributors of this volume
whose work stimulated an extremely interesting interaction with the
attendees, who in turn shall not be forgotten for being so highly
motivated and bright. It is to them that this book is dedicated.
The editors:
Gerard Chollet, Maria-Gabriella Di Benedetto, Anna Esposito, Maria
Marinaro.
Contents
Section 1 -Fundamentals of Speech Analysis and Perceptron
Articulatory Constraints on Distinctive Features
john j. Ohala ............................................................................................. 3
"Herr Miiller vivra a Taranto con i suoi colleghi austriaci"
Investigations on a fragment of Italian Phonology
Andrea Calabrese ..................................................................................... 21
Acoustic Analysis and Perception of Classes of Sounds
(vowels and consonants)
Maria Gabriella Di Benedetto and Anna Esposito ................................ 54
Speech and Voice Perception: Beyond Pattern Recognition
jean-Sylvain Lienard ............................................................................... 85
Section 2 -Speech Processing
Analysis in Automatic Recognition of Speech
Hynek Herman.5ky ........................................................................................ 115
Speech Production and Perception Models and their Applications
to Synthesis, Recognition, and Coding
Abeer Alwan, S. Narayanan, B. Strope and A. Shen ............................. 138
Section 3 -Stochastic Models for Speech
Statistical Methods for Automatic Speech Recognition
Renato de Mori ......................................................................................... 165
Statistical Modelling: from Speech Recognition to Text Translation
Hermann Ney ........................................................................................... 190
Continuous Speech Recognition with Neural Networks:
An Application to Railway Timetables Enquires
Dario Albesano, Franco Mana and Roberto Gemello ........................... 216