Phrase grouping is essential to characterize the prosody for Mandarin fluent speech. Evidence of prosodic phrase grouping has been found both in adjustments of F0 contours and temporal allocations within and across phrases. In this paper, we discuss the development of a Mandarin TTS system that integrates the prosody processing modules, such as duration modeling, F0 modeling, and break predictions. The database consists of 1292*3 syllable-tokens chopped off specially designed threephrase carrier sentences. Since temporal allocations and rhythmic structure in speech flow are carefully dealt with, the TTS system is capable of converting long paragraph text input into natural synthesized speech output.
Cite as: Pin, S., Chen, Y., Wang, H., Tseng, C. (2004) A Mandarin TTS System with an Integrated Prosodic Model. Proc. International Symposium on Chinese Spoken Language Processing, 169-172
@inproceedings{pin04_iscslp,
author={ShaoHuang Pin and Yongcheng Chen and Hsinmin Wang and Chiuyu Tseng},
title={{A Mandarin TTS System with an Integrated Prosodic Model}},
year=2004,
booktitle={Proc. International Symposium on Chinese Spoken Language Processing},
pages={169--172}
}