The SRC4VC corpus consists of smartphone-recorded speech uttered by 100 native Japanese speakers.
This corpus is designed with the aim of realizing high-quality voice conversion (VC) from end-users' degraded speech input.
The text was borrowed from existing corpora, and the voices were collected through crowdsourcing using Lancers.
In addition to the recorded voice data (48000Hz/16bit wav), the audio data restored by the unofficial implementation of Miipher (22050Hz/16bit wav) is included.
The materials may be used free of charge for research purposes, but please refrain from redistribution or use that is offensive to public order and morals.
If you wish to use this information in your paper, please cite the following paper:
Yuki Saito, Takuto Igarashi, Kentaro Seki, Shinnosuke Takamichi, Ryuuichi Yamamoto, Kentaro Tachibana, and Hiroshi Saruwatari,
"SRC4VC: Smartphone-recorded corpus for voice conversion benchmark,"
Proc. INTERSPEECH, pp. 1825--1829, Kos, Greece, Sep. 2024. (Paper)
コーパス Version 1 を公開しました (2024/02/29) / Version 1 is available online (Feb. 29, 2024)
主な開発者 (Main developers):
齋藤 佑樹 (東京大学 情報理工学系研究科) / Yuki Saito at The University of Tokyo, Japan.
五十嵐 琢斗 (東京大学 情報理工学系研究科) / Takuto Igarashi at The University of Tokyo, Japan.
関 健太郎 (東京大学 情報理工学系研究科) / Kentaro Seki at The University of Tokyo, Japan.
高道 慎之介 (東京大学 情報理工学系研究科) / Shinnosuke Takamichi at The University of Tokyo, Japan.
山本 龍一 (LINEヤフー株式会社) / Ryuichi Yamamoto at LY Corp., Japan.
橘 健太郎 (LINEヤフー株式会社) / Kentaro Tachibana at LY Corp., Japan.
猿渡 洋 (東京大学 情報理工学系研究科) / Hiroshi Saruwatari at The University of Tokyo, Japan.
謝辞 (Acknowledgements):
本研究は,LINEヤフー株式会社と東京大学 猿渡・高道研究室の共同研究プロジェクトとして実施した. / This research was conducted as a joint research project between LY Corp. and Saruwatari-Takamichi Lab. at The University of Tokyo, Japan.