This is the first book to treat two areas of speech synthesis: natural language processing and the inherent problems it presents for speech synthesis; and digital signal processing, with an emphasis on the concatenative approach. The text guides the reader through the material in a step-by-step easy-to-follow way. The book will be of interest to researchers and students in phonetics and speech communication, in both academia and industry.
Dutoit, author of the MBROLA high quality free speech synthesis project, treats the topic of speech synthesis from the perspective of two different engineering approaches. Part one concerns natural language processing and the inherent problems it presents for speech synthesis. Part two focuses on digital signal processing, with an emphasis on the concatenative approach. Topics include NLP architectures for TTS synthesis, morpho-syntactic analysis, automatic phonetization, and automatic prosody generation. Annotation c. by Book News, Inc., Portland, Or.
Thierry Dutoit graduated as an electrical engineer and PhD from the Faculté Polytechnique de Mons, Belgium, in 1988 and 1993, respectively. He is now assistant professor at the Faculté Polytechnique de Mons and consultant for AT&T Labs Research in Murray Hill, New Jersey, USA. He is the author of the widely acclaimed MBROLA high quality free speech synthesis project.
List of Figures. Foreword. Preface. 1. Introduction. Part One: From Text to its Narrow Phonetic Transcription. 2. Grammars, Inference, Parsing, and Transduction. 3. NLP Architectures for TTS Synthesis. 4. Morpho-Syntactic Analysis. 5. Automatic Phonetization. 6. Automatic Prosody Generation. Part Two: From Narrow Phonetic Transcription to Speech. 7. Synthesis Strategies. 8. Linear Prediction Synthesis. 9. Hybrid Harmonic/ Shastic Synthesis. 10. Time-Domain Algorithms. 11. Conclusions and Perspectives. Index.