FLEURS-R
收藏资源简介:
FLEURS-R数据集是由谷歌DeepMind创建的多语言语音语料库,包含102种语言的高质量并行语音和文本数据。该数据集通过应用先进的语音恢复技术,显著提升了音频的清晰度和保真度,特别适用于低资源语言的语音生成任务。数据集的创建过程中,采用了创新的语音处理管道和模型,确保了语音内容的完整性和自然性。FLEURS-R数据集主要应用于推动多语言和跨语言的语音技术研究,特别是在文本到语音和语音到语音翻译等领域。
The FLEURS-R dataset is a multilingual speech corpus created by Google DeepMind, containing high-quality parallel speech and text data across 102 languages. By applying advanced speech restoration technologies, this dataset significantly improves the clarity and fidelity of audio, making it particularly suitable for speech generation tasks in low-resource languages. Developed with an innovative speech processing pipeline and models, the dataset ensures the integrity and naturalness of speech content. The FLEURS-R dataset is primarily utilized to advance multilingual and cross-lingual speech technology research, especially in fields such as text-to-speech and speech-to-speech translation.




