Hi,
maybe you have seen ai tools that imitate voice.
i am very interested in creating a similar tool in pb. but I don't know where to start.
in your opinion, how can we take audio file (a) as input and produce audio file (b) that imitates the sound of file (a) ?
does anyone have an idea to write code in pb for this?
thanks.
Voice cloning
Re: Voice cloning
This is a very difficult task, only feasible with artificial intelligence.
You can use such online services through their APIs with PB, but doing it yourself from scratch is not a good idea and you need a lot of AI knowledge.
You can use such online services through their APIs with PB, but doing it yourself from scratch is not a good idea and you need a lot of AI knowledge.
- NicTheQuick
- Addict
- Posts: 1227
- Joined: Sun Jun 22, 2003 7:43 pm
- Location: Germany, Saarbrücken
- Contact:
Re: Voice cloning
You can not do that by yourself from scratch. And I guess it will be really annoying to do with PB anyway, even when trying to integrate already known libraries which are not easily integratable with PB. Just use Python like everyone else because it brings everything important with it, the code is already written, the models pre-trained, and the underlying libraries compatible.
The english grammar is freeware, you can use it freely - But it's not Open Source, i.e. you can not change it or publish it in altered way.