VAE-GAN

• 论文:Autoencoding beyond pixels usingALearnedSimilarityMmetric

How does a VAE-GAN work?

• We have three networks, an Encoder, a Generator, and a Discriminator.
  • The Encoder learns to map input x onto z space (latent space)
  • The Generator learns to generate x from z space
  • The Discriminator learns to discriminate whether the image being put in is real, or generated