Alleviating the over-smoothing problem in GMM-based voice conversion with discriminative training

Hsin Te Hwang, Yu Tsao, Hsin Min Wang, Yih-Ru Wang, Sin-Horng Chen

Research output: Contribution to journalConference articlepeer-review

7 Scopus citations

Abstract

In this paper, we propose a discriminative training (DT) method to alleviate the muffled sound effect caused by over smoothing in the Gaussian mixture model (GMM)-based voice conversion (VC). For the conventional GMM-based VC, we often observed a large degree of ambiguities among acoustic classes (generative classes), determined by the source feature vectors for generating the converted feature vectors, causing the "muffled sound" effect on the converted voice. The proposed DT method is applied to refine the parameters in the maximum likelihood (ML)-trained joint density GMM (JDGMM) in the training stage to reduce the ambiguities among acoustic classes (generative classes) to alleviate the muffled sound effect. Experimental results demonstrate that the DT method significantly enhances the discriminative power between acoustic classes (generative classes) in the objective evaluation and effectively alleviates the muffled sound effect in the subjective evaluation.

Original languageEnglish
Pages (from-to)3062-3066
Number of pages5
JournalProceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
StatePublished - 1 Jan 2013
Event14th Annual Conference of the International Speech Communication Association, INTERSPEECH 2013 - Lyon, France
Duration: 25 Aug 201329 Aug 2013

Keywords

  • Discriminative training
  • GMM
  • Voice conversion

Fingerprint Dive into the research topics of 'Alleviating the over-smoothing problem in GMM-based voice conversion with discriminative training'. Together they form a unique fingerprint.

Cite this