ISCA Archive PMLA 2002
ISCA Archive PMLA 2002

Pronunciation issues in text-to-speech synthesis

Paul Taylor

This talk will cover a variety of issues in text-to-speech (TTS) synthesis, specifically focusing on current issues in the design of a state of the art unit selection system. The talk will start with an overview of traditional and previous approaches to pronunciation modelling. Next, the issue whether it is best to use phonological or phonetic representations will be discussed. Issues relating to prosodic patterns of sentence stress and lexical stress will be examined. The talk will conclude with an investigation into the phonetics and phonology of regional accents, specifically on the issue of whether to represent the differences in the lexicon or the acoustics. Examples of TTS output will be used to illustrate the talk.


Cite as: Taylor, P. (2002) Pronunciation issues in text-to-speech synthesis. Proc. ITRW on Pronunciation Modeling and Lexicon Adaptation for Spoken Language Technology (PMLA 2002), 135 (Abstract only)

@inproceedings{taylor02_pmla,
  author={Paul Taylor},
  title={{Pronunciation issues in text-to-speech synthesis}},
  year=2002,
  booktitle={Proc. ITRW on Pronunciation Modeling and Lexicon Adaptation for Spoken Language Technology (PMLA 2002)},
  pages={135 (Abstract only)}
}