An Evaluation of Many-to-One Voice Conversion Algorithms with Pre-Stored Speaker Data Sets


NAIST \| Library	Advanced Search	Japanese \| English

Please use this identifier to cite or link to this item: http://hdl.handle.net/10061/8267

Title:	An Evaluation of Many-to-One Voice Conversion Algorithms with Pre-Stored Speaker Data Sets
Authors:	Daisuke Tani Yamato Ohtani Tomoki Toda Hiroshi Saruwatari Kiyohiro Shikano
Keywords:	voice conversion many-to-one VC EVC SAT speaker selection
Issue Date:	Aug-2007
Start page:	107
End page:	112
Abstract:	This paper describes an evaluation of many-to-one voice conversion (VC) algorithms converting an arbitrary speaker?s voice into a particular target speaker?s voice. These algorithms effectively generate a conversion model for a new source speaker using multiple parallel data sets of many pre-stored source speakers and the single target speaker. We conducted experimental evaluations for demonstrating the conversion performance of each of the many-to-one VC algorithms, including not only the conventional algorithms based on a speaker independent GMM and on eigenvoice conversion (EVC), but also new algorithms based on speaker selection and on EVC with speaker adaptive training (SAT). As a result, it is shown that an adaptation process of the conversion model improves significantly conversion performance, and the algorithm based on speaker selection works well even when using a very limited amount of adaptation data.
Description:	SSW6: 6th ISCA Speech Synthesis Workshop, August 22-24, 2007, Bonn, Germany.
URI:	http://hdl.handle.net/10061/8267
Text Version:	Publisher
Appears in Collections:	情報科学研究科 / Graduate School of Information Science

Files in This Item:

File	Description	Size	Format
SSW_2007_107.pdf		404.82 kB	Adobe PDF	View/Open
presentation.pdf		654.98 kB	Adobe PDF	View/Open