search

FunAudioLLM / CosyVoice

Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.
https://funaudiollm.github.io/
Apache License 2.0
4.56k stars 461 forks source link

跨语言音色克隆效果不好 #162

Open wqzh opened 1 month ago

wqzh commented 1 month ago

使用中文的一段prompt_wav （这个角色只会说中文），一段韩语的tts_txt， 希望角色能用中文的音色说出韩语。 但是合成的语音质量很差。 请问有什么操作可以改善 跨语言音色克隆 的效果？

aluminumbox commented 1 month ago

well we didn't construct enough cross lingual data. image one speaker speakers at least two language, such training data is rare.