Data for the VoiceMOS Challenge 2022

This is the public release of the data for the first VoiceMOS Challenge.  The challenge had two tracks: a main track and an out-of-domain (OOD) track.  The data for the main track is known as the BVCC dataset, and contains samples from past Blizzard Challenges, Voice Conversion Challenges, and public samples from ESPnet-TTS, along with their mean opinion score (MOS) ratings collected in one unified listening test.  Standard training/development/testing splits from the challenge are also provided.  The OOD track contains samples from the Blizzard Challenge 2019 along with their ratings from the original, separate listening test.  We also include the scoring scripts that were used for the challenge.

Samples from Blizzard Challenges may NOT be redistributed.  Blizzard samples are not included in this dataset, but the scripts to download and preprocess them are included.  Please run all of the included scripts to obtain the full dataset.

Version 2 contains a new script, get_transcripts.py, and data file for gathering the transcripts of the audio samples from the Blizzard directories after those files have been downloaded using 01_get.py.  These are in the 'main' subdirectory.

BVCC reference:Erica Cooper and Junichi Yamagishi, "How do Voices from Past Speech Synthesis Challenges Compare Today?" SSW 2021. https://arxiv.org/abs/2105.02373

The VoiceMOS Challenge:Wen-Chin Huang, Erica Cooper, Yu Tsao, Hsin-Min Wang, Tomoki Toda, Junichi Yamagishi, "The VoiceMOS Challenge 2022," submitted to Interspeech 2022.  https://arxiv.org/abs/2203.11389https://voicemos-challenge-2022.github.io

The Blizzard Challenges 2008, 2009, 2010, 2011, 2013, 2016, 2019:V. Karaiskos, S. King, R. A. Clark, and C. Mayo, "The Blizzard Challenge 2008," in Proc. Blizzard Challenge Workshop, 2008.A. W. Black, S. King, and K. Tokuda, "The Blizzard Challenge 2009," in Proc. Blizzard Challenge, 2009.S. King and V. Karaiskos, "The Blizzard Challenge 2010," 2010.S. King and V. Karaiskos, "The Blizzard Challenge 2011," 2011.S. King and V. Karaiskos, "The Blizzard Challenge 2013," 2013.S. King and V. Karaiskos, "The Blizzard Challenge 2016," 2016.Z. Wu, Z. Xie, and S. King, "The Blizzard Challenge 2019," 2019.

The Voice Conversion Challenges 2016, 2018, and 2020:T. Toda, L.-H. Chen, D. Saito, F. Villavicencio, M. Wester, Z. Wu, and J. Yamagishi, "The Voice Conversion Challenge 2016," Interspeech, 2016.J. Lorenzo-Trueba, J. Yamagishi, T. Toda, D. Saito, F. Villavicencio, T. Kinnunen, and Z. Ling, "The Voice Conversion Challenge 2018: Promoting development of parallel and nonparallel methods."Z. Yi, W.-C. Huang, X. Tian, J. Yamagishi, R. K. Das, T. Kinnunen, Z. Ling, and T. Toda, "Voice Conversion Challenge 2020 — intra-lingual semi-parallel and cross-lingual voice conversion —," in Proc. Joint Workshop for the Blizzard Challenge and Voice Conversion Challenge 2020, 2020, pp. 80–98.

ESPnet-TTS:S. Watanabe, T. Hori, S. Karita, T. Hayashi, J. Nishitoba, Y. Unno, N. Enrique Yalta Soplin, J. Heymann, M. Wiesner, N. Chen, A. Renduchintala, and T. Ochiai, "ESPnet: End-to-end speech processing toolkit," in Proceedings of Interspeech, 2018, pp. 2207–2211. [Online]. Available: http://dx.doi.org/10.21437/ Interspeech.2018- 1456

本资源为首届语音MOS挑战赛（VoiceMOS Challenge）公开发布的数据集。本次挑战赛设有两大赛道：主赛道与域外（Out-of-Domain, OOD）赛道。主赛道数据集命名为BVCC数据集，包含过往暴雪挑战赛（Blizzard Challenge）、语音转换挑战赛（Voice Conversion Challenge）的样本，以及ESPnet-TTS公开音频样本，同时附带统一听力测试中采集的平均主观得分（Mean Opinion Score, MOS）标注。本次挑战赛标准的训练/开发/测试集划分方案也已同步提供。域外赛道数据集源自2019年暴雪挑战赛，附带原独立听力测试得到的标注。本次发布还包含挑战赛使用的评分脚本。

需注意，暴雪挑战赛的样本不得进行二次分发。本数据集未直接包含暴雪样本，但提供了用于下载与预处理这些样本的脚本。请运行所有附带脚本以获取完整数据集。

V2版本新增了get_transcripts.py脚本与配套数据文件，用于在通过01_get.py下载暴雪目录下的音频样本后，提取其转录文本，相关文件存储于'main'子目录中。

BVCC相关文献：Erica Cooper与Junichi Yamagishi，"How do Voices from Past Speech Synthesis Challenges Compare Today?"，SSW 2021，https://arxiv.org/abs/2105.02373

首届语音MOS挑战赛相关文献：Wen-Chin Huang、Erica Cooper、Yu Tsao、Hsin-Min Wang、Tomoki Toda、Junichi Yamagishi，"The VoiceMOS Challenge 2022"，提交至Interspeech 2022，https://arxiv.org/abs/2203.11389，https://voicemos-challenge-2022.github.io

2008、2009、2010、2011、2013、2016、2019年暴雪挑战赛相关文献：
V. Karaiskos、S. King、R. A. Clark与C. Mayo，"The Blizzard Challenge 2008"，收录于Blizzard Challenge Workshop会议论文集，2008年。
A. W. Black、S. King与K. Tokuda，"The Blizzard Challenge 2009"，收录于Blizzard Challenge会议论文集，2009年。
S. King与V. Karaiskos，"The Blizzard Challenge 2010"，2010年。
S. King与V. Karaiskos，"The Blizzard Challenge 2011"，2011年。
S. King与V. Karaiskos，"The Blizzard Challenge 2013"，2013年。
S. King与V. Karaiskos，"The Blizzard Challenge 2016"，2016年。
Z. Wu、Z. Xie与S. King，"The Blizzard Challenge 2019"，2019年。

2016、2018、2020年语音转换挑战赛相关文献：
T. Toda、L.-H. Chen、D. Saito、F. Villavicencio、M. Wester、Z. Wu与J. Yamagishi，"The Voice Conversion Challenge 2016"，Interspeech 2016。
J. Lorenzo-Trueba、J. Yamagishi、T. Toda、D. Saito、F. Villavicencio、T. Kinnunen与Z. Ling，"The Voice Conversion Challenge 2018: Promoting development of parallel and nonparallel methods"。
Z. Yi、W.-C. Huang、X. Tian、J. Yamagishi、R. K. Das、T. Kinnunen、Z. Ling与T. Toda，"Voice Conversion Challenge 2020 — intra-lingual semi-parallel and cross-lingual voice conversion —"，收录于Blizzard Challenge与Voice Conversion Challenge联合工作坊会议论文集，2020年，第80-98页。

ESPnet-TTS相关文献：S. Watanabe、T. Hori、S. Karita、T. Hayashi、J. Nishitoba、Y. Unno、N. Enrique Yalta Soplin、J. Heymann、M. Wiesner、N. Chen、A. Renduchintala与T. Ochiai，"ESPnet: End-to-end speech processing toolkit"，收录于Interspeech 2018会议论文集，第2207-2211页，[在线]。可获取：http://dx.doi.org/10.21437/Interspeech.2018-1456