Alleviating the one-to-many mapping problem in voice conversion with context-dependent modelling

Elizabeth Godoy (1) , Olivier Rosec (1) , Thierry Chonavel (2, 3)

1 Orange Labs - Orange Labs [Belfort]
2 Lab-STICC_TB_CID_TOMS
3 SC - Département Signal et Communications

This paper addresses the "one-to-many" mapping problem in Voice Conversion (VC) by exploring source-to-target mappings in GMM-based spectral transformation. Specifically, we examine differences using source-only versus joint source/target information in the classification stage of transformation, effectively illustrating a "one-to-many effect" in the traditional acoustically-based GMM. We propose combating this effect by using phonetic information in the GMM learning and classification. We then show the success of our proposed context-dependent modeling with transformation results using an objective error criterion. Finally, we discuss implications of our work in adapting current approaches to VC.

Mots clés: Voice conversion; GMM; spectral mapping

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

Elizabeth Godoy, Olivier Rosec, Thierry Chonavel. Alleviating the one-to-many mapping problem in voice conversion with context-dependent modelling. InterSpeech 09 : 10th Annual Conference of the International Speech Communication Association, Sep 2009, Brighton, United Kingdom. ⟨hal-00498445⟩