紀伊國屋書店

基本説明

Speech processing research in Japan started in the 1940s. In the 1990s, reserach on spoken language processing became so successful that it was extended to large vocabulary, continuous speech recognition, corpus-based speech synthesis, spoken dialogue systems, multi-modal interfaces and speech-to-speech translation systems, which are all covered by this book.

Full Description

Speech processing research in Japan started in the 1940s. While it has produced high quality results, many works were described only in Japanese and were not necessarily made available to the global community. In the 1990s, research on spoken language processing became so successful that it was extended to large vocabulary, continuous speech recognition, corpus-based speech synthesis, spoken dialogue systems, multi-modal interfaces and speech-to-speech translation systems, which are all covered by this book. It is believed that most of the basic ideas and technologies behind current spoken language systems were established by 2000. This book provides a compendium of the prominent studies on such prototypes of spoken language systems developed in Japan. This book is a comprehensive introduction to the major works conducted at Japanese research institutes that are developing spoken language systems.

Contents

Large Vocabulary Continuous Speech Recognition Efforts in Japan; Speech Recognition System in NEC; IBM's Japanese Dictation System; Real-Time Transcription System for Simultaneous Subtitling of Broadcast News Programs; Modelling and Generation of Prosodic Features; Challenges for New Speech Synthesis Applications; Concept-to-Speech Conversion System for Spoken Dialogue Systems; Spoken Dialogue Systems in Japan -- Review; The Spoken Dialogue System of TUT; Spoken Dialogue System DUG-1; Speech Recognition Techniques and Applications Developed by KDD; Multi-modal Systems and Rapid-Prototyping Tools; Multi-modal Interactive Systems using Speech and Pointing Gestures; Real-time Gesture-Speech Human Interface Realised on a Personal Computer of Notebook Size; Multi-Modal Conversation Robot Participating Group Conversation; ATR-MATRIX: A Spontaneous Speech Translation System and C-STAR; Robust Distant-Talking Speech Recognition; Auditory-Like Acoustic Processing; Language Modelling and Adaptation; Spoken Document Retrieval using Speaker and Speech Recognition; Speaker Verification; Spoken Language Resources; Discourse Annotation Schemes for Japanese Task-Oriented Dialogues.