In this paper, we improve the MGE criterion by imposing a log spectral distortion (LSD) instead of the Euclidean distance to define the generation error between ...

A minimum generation error (MGE) criterion had been pro- posed to solve the issues related to maximum likelihood (ML) based HMM training in HMM-based speech ...

From the experimental results, the generation errors were reduced after the MGE-based HMM training, and the quality of synthetic speech is improved.

This paper improves a minimum generation error (MGE) based HMM training technique for HMM-based speech synthesis by directly using the original spectrum ...

Minimum generation error training with direct log spectral distortion on LSPs for HMM-based speech synthesis ; Toloosham B., Koishida K. · 2022 , citations by ...

Tokuda, “Minimum generation error training with di- rect log spectral distortion on LSPs for HMM-based speech synthesis,” in Proc. of Interspeech, pp. 577 ...

This paper utilizes global variance (GV) of the log power spectrum (LPS) derived from mel-cepstrum to improve hidden Markov model (HMM) based parametric speech ...

In this paper, an MGE training criterion based on weighted LSP distance is proposed, and this MGE training method is subjectively and objectively contrasted ...

Minimum generation error training with direct log spectral distortion on LSPs for HMM-based speech synthesisYi-Jian Wu, Keiichi Tokuda. interspeech 2008 ...

Tokuda, “Minimum generation error training with direct log spectral distortion on LSPs for HMM-based speech synthesis,” in Proc. of Interspeech, pp. 577–580 ...