Skip to content

Latest commit

 

History

History
29 lines (18 loc) · 761 Bytes

README.md

File metadata and controls

29 lines (18 loc) · 761 Bytes

Anime Face Customization

Generation of Anime faces through the latent space of a Variational Autoencoder (VAE).

Built with Pytorch, Using Danbooru2018 Dataset.

Example

Random Samples

Requirements:

  • Pytorch
  • Visdom
  • lbpcascade_animeface
  • Collection of anime-style images (Danbooru2018)

Insights:

  • When using plain VAE, balancing the KL Divergence and the MSE Loss is hard. Images tend to be blurry.
  • With RMSprop and fine tuning convergence tends to be really slow at the end

References: