malaysia-ai/Emilia-YODAS-Voice-Conversion
收藏Hugging Face2025-12-05 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/malaysia-ai/Emilia-YODAS-Voice-Conversion
下载链接
链接失效反馈官方服务：
资源简介：
Emilia-YODAS语音转换数据集，包含了德语、英语、法语、日语、韩语和中文等多种语言的语音数据。数据集通过特定的转录过滤和说话人相似度过滤，以及使用neucodec模型将音频转换为语音标记。数据集总共有5.7亿个语音标记，并且提供了每种语言的语音时长统计信息。
The Emilia-YODAS Voice Conversion dataset includes multilingual voice data such as German, English, French, Japanese, Korean, and Chinese. The dataset is processed through specific transcription filtering and speaker similarity filtering, and audio is converted into speech tokens using the neucodec model. The dataset consists of a total of 570 million speech tokens and provides statistics on the speaking duration of each language.
提供机构：
malaysia-ai
