Muutke küpsiste eelistusi

Spoken Dialogue With Computers [Kõva köide]

(McGill University, Montréal, Québec, Canada and Université d'Avignon, France)
  • Formaat: Hardback, 702 pages, kõrgus x laius: 229x152 mm, kaal: 1460 g
  • Sari: Signal Processing and Its Applications
  • Ilmumisaeg: 06-Apr-1998
  • Kirjastus: Academic Press Inc
  • ISBN-10: 0122090551
  • ISBN-13: 9780122090554
Teised raamatud teemal:
Spoken Dialogue With ComputersSuurem pilt
  • Formaat: Hardback, 702 pages, kõrgus x laius: 229x152 mm, kaal: 1460 g
  • Sari: Signal Processing and Its Applications
  • Ilmumisaeg: 06-Apr-1998
  • Kirjastus: Academic Press Inc
  • ISBN-10: 0122090551
  • ISBN-13: 9780122090554
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9780122090554.html
Märksõnad:
Presents a description of all of the components of a simplified version of the architecture of computer-based spoken-dialogue systems currently in use. Including algorithmic presentation of important components, 17 chapters follow the architecture in a step-by-step format that covers microphones, basic speech analysis, parameter transformation, acoustic and language modeling, knowledge integration, search and generation of word hypotheses, neural networks, speaker adaptation, the use of formal grammars, sentence interpretation and generation, and system architecture and applications. Annotation c. by Book News, Inc., Portland, Or.

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

Muu info

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

1. Problems and Methods for Solution 1(22) Renato De Mori 1.1. Introduction 1(2) 1.2. Problem position and methods for solution 3(4) 1.3. Book structure 7(4) 1.4. System architectures 11(3) 1.5. Historical notes 14(3) 1.6. Activities and problems in different languages for ASR 17(5) Acknowledgements 22(1)
2. Acoustic Transduction 23(46) Maurizio Omologo Piergiorgio Svaizer Renato De Mori 2.1. Introduction 24(1) 2.2. Sound propagation 24(7) 2.3. Acoustic sensors 31(7) 2.4. Microphone arrays 38(13) 2.5. Talker location 51(13) 2.6. Microphones in speech recognition 64(5)
3. Basic Speech Sounds, their Analysis and Features 69(54) Bianca Angelini Daniele Falavigna Maurizio Omologo Renato De Mori 3.1. Introduction 70(1) 3.2. The nature of speech sounds 70(14) 3.3. Speech analysis principles 84(20) 3.4. Auditory models 104(2) 3.5. Acoustic features 106(11) 3.6. Dynamic features 117(3) 3.7. Conclusions 120(3)
4. Parameter Transformation 123(18) Diego Giuliani Daniele Falavigna Renato De Mori 4.1. Introduction 123(1) 4.2. Dimensionality reduction 124(7) 4.3. Vector quantization 131(7) 4.4. Applications and trends 138(3)
5. Acoustic Modelling 141(30) Fabio Brugnara Renato De Mori 5.1. Introduction 141(1) 5.2. Hidden Markov model theory 142(19) 5.3. Entities modelled 161(2) 5.4. Parameter tying 163(3) 5.5. HMM implementation issues 166(1) 5.6. Extension and variations 167(1) 5.7. Phoneme recognition results with HMMs 168(3)
6. Training of Acoustic Models 171(28) Fabio Brugnara Renato De Mori 6.1. Introduction 171(1) 6.2. The Baum-Welch algorithm 172(2) 6.3. Maximum likelihood estimation 174(11) 6.4. Maximum a posteriori (MAP) estimation 185(5) 6.5. Generation of seed models 190(1) 6.6. Notes on the Baum-Welch algorithm 191(3) 6.7. Other training methods 194(5)
7. Language Modelling 199(32) Marcello Federico Renato De Mori 7.1. Introduction to stochastic language models 200(2) 7.2. Perplexity as a measure for LM evaluation 202(2) 7.3. Basic estimation theory 204(6) 7.4. Interpolation and backing-off LMs 210(9) 7.5. Constraint-based LMs 219(4) 7.6. Variations on n-gram LMs 223(5) 7.7. Applications of LMs 228(3)
8. Knowledge Integration 231(26) Mauro Cettolo Roberto Gretter Renato De Mori 8.1. Introduction 231(2) 8.2. Whole-word models vs subword models 233(2) 8.3. Generation of canonical pronunciation forms 235(4) 8.4. Multiple pronunciations 239(4) 8.5. Building an integrated network for search 243(5) 8.6. Static tree-based network allocation 248(7) 8.7. Conclusions 255(2)
9. Search and Generation of Word Hypotheses 257(54) Mauro Cettolo Roberto Gretter Renato De Mori 9.1. Introduction 257(1) 9.2. Basic search algorithms 258(7) 9.3. Search strategies applied to ASR 265(5) 9.4. Beam search 270(1) 9.5. Systems based on one-pass search 271(4) 9.6. Fast match for constraining search 275(4) 9.7. Multistage search 279(16) 9.8. Some recent results in dictation 295(2) 9.9. Word spotting 297(11) 9.10. Conclusions 308(3)
10. Neural Networks for Speech Recognition 311(52) Edmondo Trentin Yoshua Bengio Cesare Furlanello Renato De Mori 10.1. Introduction 312(2) 10.2. Simple linear perceptrons and the Widrow-Hoff algorithm 314(5) 10.3. Multilayer perceptrons and the backpropagation algorithm 319(4) 10.4. Important heuristics for training MLPs 323(5) 10.5. Radial basis function networks 328(4) 10.6. Relation between ANNs and statistical models 332(3) 10.7. Unsupervised learning 335(7) 10.8. ANNs for time sequences 342(5) 10.9. ANN/HMM hybrid systems 347(13) 10.10. Concluding remarks 360(3)
11. Speaker Adaptation 363(42) Diego Giuliani Renato De Mori 11.1. Introduction 363(3) 11.2. Batch, incremental and instantaneous speaker adaptation 366(4) 11.3. Acoustic-model adaptation 370(19) 11.4. Feature vector transformation approach 389(12) 11.5. Speaker clustering and model selection 401(1) 11.6. Conclusions 402(3)
12. Robust Speech Recognition 405(56) Chafic Mokbel Denis Jouvet Jean Monne Renato De Mori 12.1. Introduction 406(2) 12.2. Problem definition and solution classes 408(7) 12.3. Robust feature analysis 415(19) 12.4. Speech enhancement and channel equalization 434(16) 12.5. Adaptation of model parameters in new conditions 450(9) 12.6. Conclusions 459(2)
13. On the Use of Formal Grammars 461(24) Anna Corazza Renato De Mori 13.1. Introduction 461(2) 13.2. Context-free grammars 463(3) 13.3. Stochastic context-free grammars 466(3) 13.4. Probability computation 469(6) 13.5. Partial analyses 475(3) 13.6. Applications 478(1) 13.7. Some models for natural language processing 478(4) 13.8. Overview of grammars and parsers for ASR and ASU 482(1) 13.9. Conclusions and trends 483(2)
14. Sentence Interpretation 485(38) Roland Kuhn Renato De Mori 14.1. Introduction 486(1) 14.2. Semantic representation in computer systems 487(5) 14.3. Use of knowledge for language comprehension 492(5) 14.4. Control strategies for interpretation 497(4) 14.5. Semantic scores 501(3) 14.6. Semantics in ASU: the experience of the ATIS system 504(16) 14.7. Concluding remarks 520(3)
15. Dialogue Systems 523(40) David Sadek Renato De Mori 15.1. Introduction 524(2) 15.2. Simple dialogue models 526(1) 15.3. Basic dialogue phenomena 527(7) 15.4. Computational models for dialogue systems 534(3) 15.5. Principles of the rational interaction approach 537(12) 15.6. Domain-dependent constraints 549(2) 15.7. Dialogue and natural language 551(9) 15.8. Conclusion 560(3)
16. Sentence Generation 563(20) Christel Sorin Renato De Mori 16.1. Introduction 563(2) 16.2. Text generation from conceptual representations 565(8) 16.3. Text-to-sound conversion 573(2) 16.4. Use of prosody for speech synthesis 575(2) 16.5. Speech synthesizers 577(3) 16.6. State of the art and speech synthesis systems 580(2) 16.7. Concluding remarks 582(1)
17. System Architectures and Applications 583(28) Giuliano Antoniol Roberto Fiutem Gianni Lazzari Renato De Mori 17.1. Introduction and problem statement 583(1) 17.2. Software architectures 584(8) 17.3. ASR application architectures and their environment 592(5) 17.4. User interface 597(8) 17.5. ASR-based system usability 605(4) 17.6. Concluding remarks 609(2) Appendix A. Signal Processing 611(6) Daniele Falavigna Maurizio Omologo Piergiorgio Svaizer Renato De Mori Appendix B. Classification Trees 617(6) Roland Kuhn Renato De Mori Appendix C. Speech Corpora 623(4) Renato De Mori Roland Kuhn Appendix D. Formal Equations of Agent Behaviour 627(6) Giuliano Antoniol Roberto Fiutem Renato De Mori Bibliography 633(64) Index 697
Renato De Mori graduated in electronic engineering from Politeuiso di Tovino, Italy. He has been active in research since 1969,and is a fellow of the IEEE. For this book he brings together developers of deployed systems and has co-authored each chapter to ensure coherence and homogeneity. Renato De Mori is Professor of Computer Science at the University of Avignon, France. He has also been a Director at the School of Computer Science, McGill University, Montreal, Canada for the last ten years.