Make your own text to speech dataset with this tool.
Why should you make one? To replicate people's voices kinda like this and much more.
Fair Warning: It's way harder than you think and this will make it a little less harder
- Here is a sentdex video about voice cloning.
Install the application in your computer. You can find it in the releases section.
The dataset folder will look like(Similar to LJ speech dataset):
Destination folder:
-wavs <===== folder containing the clips
-metadata.csv <===== csv file containing the clip name and corresponding text
To Do:
I/O:
- Better Responsive UI.
- Add some way to begin from where it was left off.
- Add timeline to wavesurfer.
- Add keyboard shortcuts for the activities.
- Add yt and audio link support.
- Better Readme
Core additional features:
- Add slow mo option for playback.
- Remove silent parts from the clip.
Send me your queries @ [email protected]