CVSS

CVSS 是一个大规模的多语言到英语的语音到语音翻译 (S2ST) 语料库，涵盖从 21 种语言到英语的句子级并行 S2ST 对。 CVSS 源自 Common Voice 语音语料库和 CoVoST 2 语音到文本翻译 (ST) 语料库，通过使用最先进的 TTS 系统将 CoVoST 2 的翻译文本合成为语音

CVSS is a large-scale multilingual-to-English speech-to-speech translation (S2ST) corpus that contains sentence-level parallel S2ST pairs from 21 languages into English. Derived from the Common Voice speech corpus and the CoVoST 2 speech-to-text translation (ST) corpus, CVSS is built by synthesizing the translated text from CoVoST 2 into speech using state-of-the-art TTS systems.

搜集汇总

背景与挑战

背景概述

CVSS是一个大规模多语言到英语的语音到语音翻译语料库，包含21种语言到英语的句子级并行翻译对。该数据集基于Common Voice和CoVoST 2构建，通过先进TTS系统合成语音，由Google Research于2022年发布，采用CC BY 4.0许可证。