https://github.com/sanghoon/tf-exercise-gan

Tensorflow implementation of different GANs and their comparisions
https://github.com/sanghoon/tf-exercise-gan

began celeba dcgan evaluation gan gan-implementations gang-of-gans madgan mnist wgan

Last synced: 7 days ago
JSON representation

Tensorflow implementation of different GANs and their comparisions

• Host: GitHub
• URL: https://github.com/sanghoon/tf-exercise-gan
• Owner: sanghoon
• License: mit
• Created: 2017-05-25T08:35:25.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2017-11-13T13:11:59.000Z (almost 7 years ago)
• Last Synced: 2024-08-02T07:08:34.260Z (3 months ago)
• Topics: began, celeba, dcgan, evaluation, gan, gan-implementations, gang-of-gans, madgan, mnist, wgan
• Language: Python
• Homepage:
• Size: 9.99 MB
• Stars: 113
• Watchers: 16
• Forks: 32
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-GAN-papers - 16

README

        
# tf-exercise-gan

Tensorflow implementation of different GANs and their comparisions

## GAN implementations

- [x] **DCGAN** from 'Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks' (https://arxiv.org/abs/1511.06434)

- [x] **WGAN** from 'Wasserstein GAN' (https://arxiv.org/abs/1701.07875)

- [x] **BEGAN** from 'BEGAN: Boundary Equilibrium Generative Adversarial Networks' (https://arxiv.org/abs/1703.10717)

- [x] **MAD-GAN** from 'Multi-Agent Diverse Generative Adversarial Networks' (https://arxiv.org/abs/1704.02906)

- [x] **GoGAN** from 'Gang of GANs: Generative Adversarial Networks with Maximum Margin Ranking' (https://arxiv.org/abs/1704.04865)

- [ ] ... (To be added)

## Tasks

- [x] Impl. DCGAN, GoGAN, WGAN

- [x] Impl. BEGAN, MAD-GAN

- [x] Reproduce GANs on MNIST and CelebA datasets

- [x] Impl. training & evaluation on synthetic datasets

- [x] Add sample results

- [ ] Impl. inference-only code for GANs (may require refactoring)

- [ ] Impl. better evaluation function for real images (e.g. IvOM, energy dist., ...)

- [ ] Impl. a result logger

- [x] Compare GANs (synthetic)

- [x] Compare GANs (MNIST and CelebA dataset)

- [ ] Add quantitative comparisons

- [ ] Add more GAN implementations

## Experiments & Benchmarks

170718 / [Comparison of different GAN models on synthetic datasets](assets/170718_synthetic/report_synthetic.md)

- Done without any hyper-parameter search.

- MAD-GAN worked best in the tested datasets.

- ![MADGAN_Spiral](assets/170718_synthetic/MADGAN_SynSpiral_toydisc_toydisc_LR=0.0001_NGEN=8.gif)

170718 / [Sample results on MNIST dataset](assets/170718_mnist/samples.md)

- ![WGAN_MNIST](assets/170718_mnist/WGAN_mnist_099.png)

170809 / [Sample results on CelebA dataset](assets/170809_celeba/samples.md)

- ![BEGAN_CELEBA](assets/170809_celeba/fig_BEGAN_gen_0096.png)

## Other References