Skip to content

simple implementations of different kinds of VAE in tf.keras

License

Notifications You must be signed in to change notification settings

s-omranpour/X-VAE-keras

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

X-VAE-keras

Simple implementations of different kinds of VAE with tf.keras.

VAE models included are:

  1. simple VAE
  2. VAE with mmd loss instead of classic kl-divergence loss
  3. VQ-VAE with a PixelCNN prior (thanks to Amelie Royer)
  4. Categorical VAE (Gumbel-Softmax)
  5. VAEGAN
  6. Conditional VAE
  7. VAE with 2D latent space

P.S: Implementations were made based on the original paper of each model.
Codes were written based on tensorflow=1.15 and its keras module.
All models are tested with MNIST dataset and useful plots are included.