Tensorflow implementation of various GANs and VAEs.
Pytorch version of this repository is availabel at https://github.com/znxlwm/pytorch-generative-model-collections
https://github.com/google/compare_gan is the code that was used in the paper.
It provides IS/FID and rich experimental results for all gan-variants.
Name | Paper Link | Value Function |
---|---|---|
GAN | Arxiv | ![]() |
LSGAN | Arxiv | ![]() |
WGAN | Arxiv | ![]() |
WGAN_GP | Arxiv | ![]() |
DRAGAN | Arxiv | ![]() |
CGAN | Arxiv | ![]() |
infoGAN | Arxiv | ![]() |
ACGAN | Arxiv | ![]() |
EBGAN | Arxiv | ![]() |
BEGAN | Arxiv | ![]() |
Network architecture of generator and discriminator is the exaclty sames as in infoGAN paper.
For fair comparison of core ideas in all gan variants, all implementations for network architecture are kept same except EBGAN and BEGAN. Small modification is made for EBGAN/BEGAN, since those adopt auto-encoder strucutre for discriminator. But I tried to keep the capacity of discirminator.
The following results can be reproduced with command:
python main.py --dataset mnist --gan_type <TYPE> --epoch 25 --batch_size 64
All results are randomly sampled.
Name | Epoch 2 | Epoch 10 | Epoch 25 |
---|---|---|---|
GAN | ![]() |
![]() |
![]() |
LSGAN | ![]() |
![]() |
![]() |
WGAN | ![]() |
![]() |
![]() |
WGAN_GP | ![]() |
![]() |
![]() |
DRAGAN | ![]() |
![]() |
![]() |
EBGAN | ![]() |
![]() |
![]() |
BEGAN | ![]() |
![]() |
![]() |
Each row has the same noise vector and each column has the same label condition.
Name | Epoch 1 | Epoch 10 | Epoch 25 |
---|---|---|---|
CGAN | ![]() |
![]() |
![]() |
ACGAN | ![]() |
![]() |
![]() |
infoGAN | ![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Comments on network architecture in mnist are also applied to here.
Fashion-mnist is a recently proposed dataset consisting of a training set of 60,000 examples and a test set of 10,000 examples. Each example is a 28x28 grayscale image, associated with a label from 10 classes. (T-shirt/top, Trouser, Pullover, Dress, Coat, Sandal, Shirt, Sneaker, Bag, Ankle boot)
The following results can be reproduced with command:
python main.py --dataset fashion-mnist --gan_type <TYPE> --epoch 40 --batch_size 64
All results are randomly sampled.
Name | Epoch 1 | Epoch 20 | Epoch 40 |
---|---|---|---|
GAN | ![]() |
![]() |
![]() |
LSGAN | ![]() |
![]() |
![]() |
WGAN | ![]() |
![]() |
![]() |
WGAN_GP | ![]() |
![]() |
![]() |
DRAGAN | ![]() |
![]() |
![]() |
EBGAN | ![]() |
![]() |
![]() |
BEGAN | ![]() |
![]() |
![]() |
Each row has the same noise vector and each column has the same label condition.
Name | Epoch 1 | Epoch 20 | Epoch 40 |
---|---|---|---|
CGAN | ![]() |
![]() |
![]() |
ACGAN | ![]() |
![]() |
![]() |
infoGAN | ![]() |
![]() |
![]() |
Without hyper-parameter tuning from mnist-version, ACGAN/infoGAN does not work well as compared with CGAN.
ACGAN tends to fall into mode-collapse.
infoGAN tends to ignore noise-vector. It results in that various style within the same class can not be represented.
![]() |
![]() |
![]() |
![]() |
(to be added)
Name | Paper Link | Loss Function |
---|---|---|
VAE | Arxiv | ![]() |
CVAE | Arxiv | ![]() |
DVAE | Arxiv | (to be added) |
AAE | Arxiv | (to be added) |
Network architecture of decoder(generator) and encoder(discriminator) is the exaclty sames as in infoGAN paper. The number of output nodes in encoder is different. (2x z_dim for VAE, 1 fo