likethis85/wav2lip_288x288

This is a 288x288 wav2lip model version.

The original repo: https://github.com/Rudrabha/Wav2Lip Some Features I will implement here

• input size 288x288
• PRelu
• LeakyRelu
• Gradient penalty
• Wasserstein Loss
• SAM-UNet: https://github.com/1343744768/Multiattention-UNet

I trained my own model on AVSPEECH dataset and then transfer learning with my private dataset.

Citing

To cite this repository:

@misc{Wav2Lip,
  author={Rudrabha},
  title={Wav2Lip: Accurately Lip-syncing Videos In The Wild},
  year={2020},
  url={https://github.com/Rudrabha/Wav2Lip}
}