Multilingual Speech Processing
Tanja Schultz, Katrin Kirchhoff
Academic Press, Jun 12, 2006 - Technology & Engineering - 536 pages
Tanja Schultz and Katrin Kirchhoff have compiled a comprehensive overview of speech processing from a multilingual perspective. By taking this all-inclusive approach to speech processing, the editors have included theories, algorithms, and techniques that are required to support spoken input and output in a large variety of languages. This book presents a comprehensive introduction to research problems and solutions, both from a theoretical as well as a practical perspective, and highlights technology that incorporates the increasing necessity for multilingual applications in our global community.
Current challenges of speech processing and the feasibility of sharing data and system components across different languages guide contributors in their discussions of trends, prognoses and open research issues. This includes automatic speech recognition and speech synthesis, but also speech-to-speech translation, dialog systems, automatic language identification, and handling non-native speech. The book is complemented by an overview of multilingual resources, important research trends, and actual speech processing systems that are being deployed in multilingual human-human and human-machine interfaces.
Researchers and developers in industry and academia with different backgrounds but a common interest in multilingual speech processing will find an excellent overview of research problems and solutions detailed from theoretical and practical perspectives.
* State-of-the-art research with a global perspective by authors from the USA, Asia, Europe, and South Africa
* The only comprehensive introduction to multilingual speech processing currently available
* Detailed presentation of technological advances integral to security, financial, cellular and commercial applications
What people are saying - Write a review
We haven't found any reviews in the usual places.
Other editions - View all
accent accuracy acoustic models adaptation algorithm alignment applications approach Arabic articulatory features automatic bigram bilingual BTEC components Computer context corpora corpus coverage database decision tree decoding dialect dialog systems domain English error rate evaluation example Figure French German grammar graphemes guage input interaction interfaces interlingua ISIS Japanese language identification language independent language model Language Processing language resources language technology lexical lexicon linguistic LVCSR machine translation Mandarin Chinese monolingual morphemes morphological multilingual multilingual speech N-gram native speakers node non-native speakers non-native speech output parameters performance phonemes phonotactic phrase pronunciation dictionary prosodic recognizer Schultz scores segmentation semantic sentence shared Spanish speech processing speech synthesis speech translation spoken language statistical Table target language task techniques test set tion training data transcription translation system units utterances vocabulary voice browser vowels VXML Waibel word error rate word lists writing systems
Page xxii - Jeffrey Mogul received an SB from the Massachusetts Institute of Technology in 1979, and his MS and Ph.D. degrees from Stanford University in 1980 and 1986. Since 1986, he has been a researcher at Digital's Western Research Laboratory, working on network and operating systems issues for highperformance computer systems.
Page 460 - On the approximate realization of continuous mapping by neural networks, Neural Networks, 2, 183-193.
Page 12 - Simultaneous and X. Affricates and double articulations can be represented by two symbols joined by a tie bar if necessary.