UDAR
System of speech synthesis
| First version: | June 1994 | Last version: | 7.5 1999 |
| Application: | ||||
| Used as a
part of speech synthesis systems, and separely for
phonetic studies. Some components such as Phonetic Automatic Transcriber may be used for editing books for children and foreigners. |
||||
| Description: | ||||
| UDAR is a linguistic processor preparing texts for speech synthesis. It performs: | ||||
|
||||
| The version UDAR 4.51 (1996) was made for RUSVOX and oriented to the diphonic synthesis. However, PSOLA technology is sufficient to reproduce normative pronunciation but is incapable to keep individual speakers characteristics. Thats why the development of allophonic synthesis was started. For that UDAR (version 6.1, 1997 and higher) algorithms had to be essentially modified, the dictionary grew from 90,000 words to 160,000, and the number of acoustical contours increased from 7 to 26. | ||||
| Languages processed: | ||||
| Russian | ||||
| Type of processing: | ||||
| Morphological analysis, resolution of
contextual ambiguities, words accentuation by means of a
morpho-accentologic dictionary. Marking pauses and intonation on the basis of selection of syntagmes and syntactic groups. |
||||
| Hardware/software requirements: | ||||
| IBM PC compatible, Windows 95/98 | ||||
| Developer team: | Contact person: | |||
| Department for phonetics and methods of teaching foreign languages, SPbGU | Igor' V. Zharkov Phone: (+7 812) 328 95 65 E-mail: igor@IZ2803.spb.edu, paul@phonet.lang.pu.ru |
|||