期刊名称：Journal of Computing and Information Technology
印刷版ISSN：1330-1136
电子版ISSN：1846-3908
出版年度：1997
卷号：5
期号：1
页码：11-19
语种：English
出版社：SRCE - Sveučilišni računski centar
摘要：A text-to-speech system, capable of synthesising continuous Slovenian speech from an arbitrary input text is described. The text-to-speech system is based on the concatenation of basic speech units, diphones, using the TD-PSOLA technique, and no special hardware is required. The input text is transformed into its spoken equivalent by a series of the modules. The modules, constituting the text-to-speech system are described in detail. Special attention is paid to segmental duration determination, where the effect of speaking rate on phone duration is widely studied. Finally, the results of output speech quality assessment are given in terms of acceptability and intelligibility.
关键词：text-to-speech synthesis; diphone concatenation; prosody modelling; grapheme-to-phoneme conversion; Slovenian language
