Thai Speech Synthesis based on HMM.pptVIP

Outlines Objectives To implement an HMM-based speech synthesis system for Thai language with the highest correctness of tone. Study of Thai tones Characteristics of Thai tones Study of Thai tones Characteristics of Thai tones Study of Thai tones Categorizations of Thai tones HMM-based speech synthesizer Tree-based context clustering Construction of contextual factors Tree-based context clustering Design of decision-tree structures Tree-based context clustering Design of decision-tree structures Tree-based context clustering Design of 8 context clustering styles (a)-(h) Sentence structure analysis Word structure analysis Full context labeling Construction of question set for context clustering Feature extraction Experiments Evaluation of overall tone correctness Experiments Evaluation of overall tone correctness Experiments Evaluation of overall tone correctness Experiments Evaluation of tone correctness for each tone type Experiments Evaluation of syllable duration distortion Examples of synthesized speech Conclusions Objectives Study of Thai tones Construction of contextual factors Design of decision-tree structures Design of context clustering styles Characteristics of Thai tones Categorizations of Thai tones Tree-based context clustering Evaluation of overall tone correctness Evaluation of tone correctness for each tone type Evaluation of syllable duration distortion Experiments Conclusions Syllable Structure [Nakasakul2002] Thai : Tonal Language ??? r-a-k^-3 (love) ?????? r-va-j^-2 (always) ????? khr-e-ng^-2 (strict) ?????? khr-ia-t^-2 (stress) ??? l-x-3 (and) ????? phl-iia-0 (exhausted) ???? s-iia-4 (spoil) ??? pr-i-1 (break) F0 contours of Standard Thai Tones (normalized duration) [Luksaneeyanawin1992] ????? Middle(0) ??? Low(1) ?? Falling(2) ??? High(3) ????? Rising(4) Abramson divided the tones into two groups: static group dynamic group According to the final trend of contours: upward trend group downward trend group Phoneme