15th International Congress of Phonetic Sciences (ICPhS-15)

Barcelona, Spain
August 3-9, 2003

Modeling Rhythmic Variation in Thai and its Application to Speech Synthesis

Hansjörg Mixdorff (1), Sudaporn Luksaneeyanawin (2), Patavee Charnvivit (2), Nuttakorn Thubthong (2)

(1) Berlin University of Applied Sciences, Germany
(2) Chulalongkorn University, Thailand

This study concerns a preliminary experiment on modeling the duration of Thai syllables. It is based on a corpus of minimal pairs of sentences only differing as to their stress patterns. Following a factor analysis of syllabic durations in the corpus a simple duration model was developed. This model was used for re-synthesizing the utterances by manipulating speech from a Thai TTS system by adjusting syllable durations and monotozing the F0 contour. A perception experiment was conducted with respect to the discrimination of members in each minimal pair. Although the results show that natural utterances are identified more easily, the synthetic utterances were at least correctly identified well above chance level. In prosodically ambiguous cases subjects tend to select the semantically 'more plausible' interpretation.

Full Paper

Bibliographic reference.  Mixdorff, Hansjörg / Luksaneeyanawin, Sudaporn / Charnvivit, Patavee / Thubthong, Nuttakorn (2003): "Modeling rhythmic variation in Thai and its application to speech synthesis", In ICPhS-15, 2457-2460.