Prosody in Speech Understanding Systems

Ralf Kompe (eds.)3540635807, 9783540635802

Speech technology, the automatic processing of (spontaneously) spoken language, is now known to be technically feasible. It will become the major tool for handling the confusion of languages with applications including dictation systems, information retrieval by spoken dialog, and speech-to-speech translation. The book gives a throrough account of prosodic phenomena. The author presents in detail the mathematical and comnputational background of the algorithms and statistical models used and develops algorithms enabling the exploitation of prosodic information on various levels of speech understanding, such as syntax, semantics, dialog, and translation. Then he studies the integration of these algorithms in the speech-to-speech translation system VERBMOBIL and in the dialog system EVAR and analyzes the results.

---

Table of contents :
Introduction….Pages 1-20
Basic approaches….Pages 21-65
The baseline ASU systems….Pages 67-93
Prosody….Pages 95-133
Prosodic labeling and speech corpora….Pages 135-186
Preprocessing and classification….Pages 187-209
Prosodic phrase models….Pages 211-235
Intergration of the prosodic attributes in ASU….Pages 237-283
Future work….Pages 285-290
Summary….Pages 291-297