论文信息 - CMU Blizzard 2008: Optimally using a large database for unit selection synthesis.

CMU Blizzard 2008: Optimally using a large database for unit selection synthesis.

This paper describes CMU’s entry for the Blizzard Challenge 2008. Our eventual system consisted of a fairly conventional layered cluster based unit selection system using the most predictable subset of the whole UK speech databases. This paper describes the methods we used to find the most reliable subset and the techniques used to optimize the selection. An additional technique that was used was to automatically detect “hard” text and modify the phrasing algorithm accordingly. Although this technique was targeted at SUS utterances it was in place for all utterances. CMU’s entry is letter M in the results.

[1] Susan Fitt,et al. Synthesis of regional English using a keyword lexicon , 1999, EUROSPEECH.

[2] Keiichi Tokuda,et al. The blizzard challenge - 2005: evaluating corpus-based speech synthesis on common datasets , 2005, INTERSPEECH.

[3] Kishore Prahallad,et al. Significance of early tagged contextual graphemes in grapheme based speech synthesis and recognition systems , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[4] Alan W. Black,et al. CLUSTERGEN: a statistical parametric synthesizer using trajectory modeling , 2006, INTERSPEECH.

[5] D. Hunter. MM algorithms for generalized Bradley-Terry models , 2003 .

[6] Kishore Prahallad,et al. Automatic building of synthetic voices from large multi-paragraph speech databases , 2007, INTERSPEECH.

[7] Alan W. Black,et al. Intelligibility of machine translation output in speech synthesis , 2006, INTERSPEECH.

[8] Paul Taylor,et al. Automatically clustering similar units for unit selection in speech synthesis , 1997, EUROSPEECH.