The improvement of Text-to-Speech (TTS) synthesizers' speech quality and naturalness is a continuous concern of researchers worldwide. The present paper gives a brief introduction of several previous Hansori TTS systems and is introducing our approach on experimenting, adopting and implementing corpus-based techniques for the system. We are focusing on corpus selection, on the optimal unit search criteria, on the missing unit (triphone unit) replacement handling, and we present some useful development tool options included in the trial version of the system.
Cite as: Ferencz, A., Choi, S.-W., Song, H.-E., Koo, M.-W. (2001) Hansori 2001 - corpus-based implementation of the Korean hansori text-to-speech synthesizer. Proc. 7th European Conference on Speech Communication and Technology (Eurospeech 2001), 841-844, doi: 10.21437/Eurospeech.2001-258
@inproceedings{ferencz01_eurospeech,
author={Attila Ferencz and Sung-Woo Choi and Ho-Eun Song and Myoung-Wan Koo},
title={{Hansori 2001 - corpus-based implementation of the Korean hansori text-to-speech synthesizer}},
year=2001,
booktitle={Proc. 7th European Conference on Speech Communication and Technology (Eurospeech 2001)},
pages={841--844},
doi={10.21437/Eurospeech.2001-258}
}