Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/mingtaoguo/dcgan_wgan_wgan-gp_lsgan_sngan_rsgan_began_acgan_pggan_tensorflow

Implementation of some different variants of GANs by tensorflow, Train the GAN in Google Cloud Colab, DCGAN, WGAN, WGAN-GP, LSGAN, SNGAN, RSGAN, RaSGAN, BEGAN, ACGAN, PGGAN, pix2pix, BigGAN
https://github.com/mingtaoguo/dcgan_wgan_wgan-gp_lsgan_sngan_rsgan_began_acgan_pggan_tensorflow

acgan began colab dcgan face-generative generative-adversarial-network pggan rasgan rsgan sngan tensorflow wgan wgan-gp

Last synced: 1 day ago
JSON representation

Implementation of some different variants of GANs by tensorflow, Train the GAN in Google Cloud Colab, DCGAN, WGAN, WGAN-GP, LSGAN, SNGAN, RSGAN, RaSGAN, BEGAN, ACGAN, PGGAN, pix2pix, BigGAN

Host: GitHub
URL: https://github.com/mingtaoguo/dcgan_wgan_wgan-gp_lsgan_sngan_rsgan_began_acgan_pggan_tensorflow
Owner: MingtaoGuo
License: mit
Created: 2018-07-03T09:31:32.000Z (over 6 years ago)
Default Branch: master
Last Pushed: 2022-08-20T01:47:15.000Z (over 2 years ago)
Last Synced: 2024-08-06T10:12:31.815Z (5 months ago)
Topics: acgan, began, colab, dcgan, face-generative, generative-adversarial-network, pggan, rasgan, rsgan, sngan, tensorflow, wgan, wgan-gp
Language: Python
Homepage:
Size: 2.13 MB
Stars: 222
Watchers: 7
Forks: 57
Open Issues: 4
Metadata Files:
- Readme: README.md
- License: LICENSE

Awesome Lists containing this project

README

        # DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN_RSGAN_BEGAN_ACGAN_PGGAN_pix2pix_BigGAN

Implementation of some different variants of GANs 

## Introduction

--------------

This code is mainly implement some basic GANs about 'DCGAN', 'WGAN', 'WGAN-GP', 'LSGAN', 'SNGAN', 'RSGAN'&'RaSGAN', 'BEGAN', 'ACGAN', 'PGGAN', 'pix2pix', 'BigGAN'. 

More details of these GANs, please see follow papers:

1. DCGAN: [Unsupervised representation learning with deep convolutional generative adversarial networks](https://arxiv.org/pdf/1511.06434.pdf%C3%AF%C2%BC%E2%80%B0)

2. WGAN: [Wasserstein gan](https://arxiv.org/pdf/1701.07875.pdf?__hstc=200028081.1bb630f9cde2cb5f07430159d50a3c91.1524009600081.1524009600082.1524009600083.1&__hssc=200028081.1.1524009600084&__hsfp=1773666937)

3. WGAN-GP: [Improved training of wasserstein gans](https://papers.nips.cc/paper/7159-improved-training-of-wasserstein-gans.pdf)

4. LSGAN: [Least Squares Generative Adversarial Networks](http://openaccess.thecvf.com/content_ICCV_2017/papers/Mao_Least_Squares_Generative_ICCV_2017_paper.pdf)

5. SNGAN: [Spectral normalization for generative adversarial networks](https://arxiv.org/pdf/1802.05957.pdf)

6. RSGAN&RaSGAN: [The relativistic discriminator: a key element missing from standard GAN](https://arxiv.org/abs/1807.00734)

7. BEGAN:[BEGAN: Boundary Equilibrium Generative Adversarial Networks](https://arxiv.org/pdf/1703.10717.pdf)

8. ACGAN: [Conditional Image Synthesis With Auxiliary Classifier GANs](https://arxiv.org/pdf/1610.09585.pdf)

9. PGGAN: [Progressive Growing of GANs for Improved Quality, Stability, and Variation](https://arxiv.org/pdf/1710.10196)

10. pix2pix: [Image-to-Image Translation with Conditional Adversarial Networks](https://arxiv.org/pdf/1611.07004.pdf)

11. BigGAN: [Large Scale GAN Training for High Fidelity Natural Image Synthesis](https://arxiv.org/pdf/1809.11096.pdf) [[Code]](https://github.com/MingtaoGuo/BigGAN-tensorflow)

## Attention

If your computer don't have GPU to accelerate the training process, please click [Google Cloud Colab](https://colab.research.google.com/drive/1BKGcw58kOQc4mxxm4VbAJ6BX-DEzZtgE) to train the GANs.

## How to use 

Firstly, you should download the data 'facedata.mat' from [Baidu Drive](https://pan.baidu.com/s/12fcKytGOW222bS5BccteYw) or [Google Drive](https://drive.google.com/open?id=1ROGET9rA5WAdU3C8Lfs5mxg5ufLD2uCO), then put the file 'facedata.mat' into the folder 'TrainingSet'.

## Requirements

1. python3.5

2. tensorflow1.4.0

3. pillow

4. scipy

5. numpy

Results of this code

--------------------

This result is using DCGAN trained about 8000 iterations.



Compare LSGAN, WGAN, WGAN-GP, SNGAN, RSGAN of different iteration

-----------------------------------------------------

![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/GAN.jpg)

Convergence of BEGAN

------------------------

![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/BEGAN.jpg)

![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/BEGAN_converge.jpg)

ACGAN for face generating

--------------------------

dataset: download address: [Baidu Drive](https://pan.baidu.com/s/1QZ2cra5Yu-2fcQx5dH7WiQ ) password: 5egd

|Fixed label, change noise slightly|Fixed noise, change label slightly|

|-|-|

|![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/fixed_label.jpg)|![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/fixed_noise.jpg)|

PGGAN for face generating

----------------------------

![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/pggan.jpg)

SNGAN for cifar-10

--------------------

|D_loss|G_loss|results|

|-|-|-|

|![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/sngan_d_loss.jpg)|![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/sngan_g_loss.jpg)|![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/sngan_cifar.jpg)|

Pix2Pix

-----------

Dataset: Google maps download address: [http://efrosgans.eecs.berkeley.edu/pix2pix/datasets/maps.tar.gz](http://efrosgans.eecs.berkeley.edu/pix2pix/datasets/maps.tar.gz)

Edges2Shoes download address: [http://efrosgans.eecs.berkeley.edu/pix2pix/datasets/edges2shoes.tar.gz](http://efrosgans.eecs.berkeley.edu/pix2pix/datasets/edges2shoes.tar.gz)

![](https://github.com/MingtaoGuo/DCGAN_LSGAN_WGAN_WGAN-GP_SNGAN/raw/master/Image/pix2pix.jpg)