Skip to content

This implementation builded for VCC2020. Using GAN based VAE model. For non-parallel data in Tensorflow.

Notifications You must be signed in to change notification settings

Lukelluke/GAN-VAE_VoiceConversion

Repository files navigation

GAN-VAE_VoiceConversion

This implementation builded for VCC2020. Using GAN based VAE model.

Voice Conversion using GAN based VAE model, which represent as the Generator in GAN.

We use the log(mel) feature for training. And the style of GAN is total raw way, without any

transform as WGAN,SEGAN,MelGAN etc.

As for the result, though it perform better than before raw VAE model, still not able to show perfect

in converting task.

The most important is not the result, but the polished coding way & enhanced coding ability.

Looking forward for better implemention for Voice Conversion task.

About

This implementation builded for VCC2020. Using GAN based VAE model. For non-parallel data in Tensorflow.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages