Skip to content

Latest commit

 

History

History
3 lines (3 loc) · 304 Bytes

README.md

File metadata and controls

3 lines (3 loc) · 304 Bytes

Disentangled representation of vocal data

We aim at discovering directions in an autoencodeur's/GAN's latent space, that have concrete meaning. We use the dataset from Mozilla's Common Voice project (https://commonvoice.mozilla.org), that is a reliable source of vocal data and provide some labeling.