Skip to content

sapphire008/VoiceEraser

Repository files navigation

VoiceEraser

VoiceEraser is a deep learning pipeline that removes human voice from sound clips. The resulting clips can be used for re-commenting (e.g. in the case of sports videos) or as sound effects.

VoiceEraser implements a version of 1D convolutional U-Net architecture to remove voice from audio clips.

Prepare the training data by runnning "Data Preprocessing" jupyter notebook. Then train using "VoiceEraser Training Pipeline" notebook.

The model is trained on the merge of BBC Sound Effects and librivox data.

About

Erasing voice from audios to make sound effects

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published