search

auspicious3000 / autovc

AutoVC: Zero-Shot Voice Style Transfer with Only Autoencoder Loss
https://arxiv.org/abs/1905.05879
MIT License
983 stars 207 forks source link

How content loss is implemented? #62

Closed ahrzb closed 3 years ago

ahrzb commented 3 years ago

In the paper there is a content_loss which is computed based on the speaker encoder on reconstructed audio, but I don't find any mention of that in this implementation, am I missing something?

auspicious3000 commented 3 years ago

https://github.com/auspicious3000/autovc/blob/master/solver_encoder.py#L95

ahrzb commented 3 years ago

Thanks, I was missing that entirely :see_no_evil: