Sciweavers

Share
ICASSP
2009
IEEE

Using a pitch-synchronous residual codebook for hybrid HMM/frame selection speech synthesis

10 years 4 months ago
Using a pitch-synchronous residual codebook for hybrid HMM/frame selection speech synthesis
This paper proposes a method to improve the quality delivered by statistical parametric speech synthesizers. For this, we use a codebook of pitch-synchronous residual frames, so as to construct a more realistic source signal. First a limited codebook of typical excitations is built from some training database. During the synthesis part, HMMs are used to generate filter and source coefficients. The latter coefficients contain both the pitch and a compact representation of target residual frames. The source signal is obtained by concatenating excitation frames picked up from the codebook, based on a selection criterion and taking target residual coefficients as input. Subjective results show a relevant improvement compared to the basic technique.
Thomas Drugman, Alexis Moinet, Thierry Dutoit, Geo
Added 21 May 2010
Updated 21 May 2010
Type Conference
Year 2009
Where ICASSP
Authors Thomas Drugman, Alexis Moinet, Thierry Dutoit, Geoffrey Wilfart
Comments (0)
books