A comprehensive reference on the exciting growth area of spoken dialogs with computers, this text describes the components of a computer-based spoken dialog system, and will prove invaluable to researchers in industry and academia working on speech communication systems and for applications developers. This state-of-the-art book reviews the complete chain from microphone to speech synthesis. It provides methods, models, and algorithms for building a working system. Renato De Mori is coauthor of each chapter ensuring coherence and homogeneity throughout the text.
Spoken Dialogs with Computers covers in detail: transducers and microphone arrays, speech analysis and transformation, acoustic modeling and model training, language modeling, and knowledge integration for automatic speech recognition (ASR). The book also presents generation of word hypotheses, speaker adaptation, robustness and telephone application, use of syntactic and semantic knowledge, speech interpretation and dialog strategies, speech generation, and software system architectures for practical implementation.
Key Features
* All the necessary methods and models are provided for building a working systems and there is clear algorithmic presentation of the important components
* A section on automatic interpretation allows the building of a database query system in spoken language
* The book will be invaluable to researchers in industry and academia working on speech communication systems and for application developers in industry

It is with great pleasure that I present this third volume of the series "Advanced Applications in Pattern Recognition." It represents the summary of many man- (and woman-) years of effort in the field of speech recognition by tne author's former team at the University of Turin. It combines the best results in fuzzy-set theory and artificial intelligence to point the way to definitive solutions to the speech-recognition problem. It is my hope that it will become a classic work in this field. I take this opportunity to extend my thanks and appreciation to Sy Marchand, Plenum's Senior Editor responsible for overseeing this series, and to Susan Lee and Jo Winton, who had the monumental task of preparing the camera-ready master sheets for publication. Morton Nadler General Editor vii PREFACE Si parva licet componere magnis Virgil, Georgics, 4,176 (37-30 B.C.) The work reported in this book results from years of research oriented toward the goal of making an experimental model capable of understanding spoken sentences of a natural language. This is, of course, a modest attempt compared to the complexity of the functions performed by the human brain. A method is introduced for conce1v1ng modules performing perceptual tasks and for combining them in a speech understanding system.

It is an established tradition that researchers from many countries get together on the average every three years for a two week Advanced Studies Institute on Automatic Speech Recognition and Synthesis. According to ASI policies the Institute is financed by NATO. This book contains the texts of lectures and papers contributed by the attendees of the ASI which was held July 2 - 14, 1984, at Bonas, Gers, France. Focussed on New Systems and Architectures for Automatic Speech Recognition and Synthesis, this book is divided into 4 parts: (a) Review of ba8ic algorithm8 (b) SY8tem architecture and VLSI for automatic Speech (c) Software 8Y8tem8 for automatic 8peech recognition, (d) Speech 8ynthe8i8 and phonetic8. Due to the international nature of the Institute, the readers will find in this book different styles, different points of view and applications to different languages. This reflects also some characteristics of the International Association for Pattern Recognition ( APR) whose technical committee on Speech Recognition has organized this ASI. Proposed contributions have been reviewed by an Editorial Committee composed of W. Ainsworth (Kent), R. Bisiani (Pittsburgh), J. P. Haton (Nancy), W. Hess (Munich), J. L. Houle (Montreal), P. Laface (Turin), R. Moore (Malvern), H. Niemann (Erlangen) and J. Ohala (Berkeley). Typesetting of the book was performed using SYMSET facilities developed entirely by the Department of Computer Science at Concordia University. Special thanks are due to L. Lam, H. Monkiewicz and L. Thiel.

The book collects the contributions to the NATO Advanced Study Institute on "Speech Recognition and Understanding: Recent Advances, Trends and Applications", held in Cetraro, Italy, during the first two weeks of July 1990. This Institute focused on three topics that are considered of particular interest and rich of i'p.novation by researchers in the fields of speech recognition and understanding: Advances in Hidden Markov modeling, connectionist approaches to speech and language modeling, and linguistic processing including language and dialogue modeling. The purpose of any ASI is that of encouraging scientific communications between researchers of NATO countries through advanced tutorials and presentations: excellent tutorials were offered by invited speakers that present in this book 15 papers which sum marize or detail the topics covered in their lectures. The lectures were complemented by discussions, panel sections and by the presentation of related works carried on by some of the attending researchers: these presentations have been collected in 42 short contributions to the Proceedings. This volume, that the reader can find useful for an overview, although incomplete, of the state of the art in speech understanding, is divided into 6 Parts.