Project Features
Voice timbre cloning and voice style transfer
This is not TTS text-to-speech, but rather directly converting songs or speech into your own voice (unrelated to the text content).
You can train your own voice model.
It can be used as a digital human for YouTube videos or for other video creation, supporting multilingual content.
Project address
GitHub address:
This project was last updated in 2024, meaning it hasn’t been maintained for almost a year or two.
Therefore, you should consider whether to use it.
One advantage is that there are one-click pre-packaged versions available online; you can use it directly after decompression (Windows client).
It includes a web UI, instead of a command-line terminal.
Tips
Just a reminder, this is not TTS.
For text-to-speech conversion, please try other projects.