In this article-series we are reviewing the most fundamental works of Generative Adversarial Networks in Computer Vision. We start from the very beginning from concepts such as generative learning, adversarial learning. We provide some code and illustrations for educational purposes. The goal is to focus on the intuition of the models, by tackling the multiple problems that arise when training a GAN. We have thoroughly analyzed more than 20 papers in 6 different articles in a chronological order. We will continue to update the GAN series, based on the newer publications or older ones that we skipped. We do hope that this series will provide you a big overview of the field, so that you will not need to read all the literature by yourself, independent of your background on GANs.
Update 07/2020: free ebook is realesed in the AI summer website
Link to the article: Introduction to generative learning - part 1
First Author | Published | Title | Code | Conference/Journal |
---|---|---|---|---|
Ian J. Goodfellow | 10 Jun 2014 | Generative Adversarial Networks | PyTorch | NIPS |
Mehdi Mirza | 6 Nov 2014 | Conditional Generative Adversarial Nets | Tensorflow | arXiv |
Alec Radford | 19 Nov 2015 | Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks | Tensorflow | arXiv |
Xi Chen | 12 Jun 2016 | InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets | Tensorflow | NIPS |
Tim Salimans | 10 Jun 2016 | Improved Techniques for Training GANs | Tensorflow | NIPS |
Link to the article: Conditional image synthesis and 3D object generation - part 2
First Author | Published | Title | Code | Conference/Journal |
---|---|---|---|---|
Augustus Odena | 30 Oct 2016 | Conditional Image Synthesis With Auxiliary Classifier GANs | Keras TF PyTorch | arXiv |
Jiajun Wu | 24 Oct 2016 | Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling | PyTorch | NIPS |
Zinan Lin | 12 Dec 2017 | PacGAN: The power of two samples in generative adversarial networks | TF | NIPS |
Phillip Isola | 21 Nov 2016 | Image-to-Image Translation with Conditional Adversarial Networks | TF PyTorch | arXiv |
Jun-Yan Zhu | 30 Mar 2017 | Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks | TF PyTorch | arXiv |
Improved training with Wasserstein distance, game theory control and progressively growing schemes (part3)
Link to the article: Improved training with Wasserstein distance, game theory control and progressively growing schemes - part3
First Author | Published | Title | Code | Conference/Journal |
---|---|---|---|---|
Martin Arjovsky | 26 Jan 2017 | Wasserstein GAN | TF PyTorch | PMLR |
David Berthelot | 31 Mar 2017 | BEGAN: Boundary Equilibrium Generative Adversarial Networks | TF PyTorch | NIPS |
Tero Karras | 27 Oct 2017 | Progressive Growing of GANs for Improved Quality, Stability, and Variation | TF PyTorch | ICLR |
Link to the article: 2K image and video synthesis, and large-scale class-conditional image generation - part4
First Author | Published | Title | Code | Conference/Journal |
---|---|---|---|---|
Ting-Chun Wang | 30 Nov 2017 | High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs | PyTorch | CVPR |
Ting-Chun Wang | 20 Aug 2018 | Video-to-Video Synthesis | PyTorch | NIPS |
Andrew Brock | 28 Sep 2018 | Large Scale GAN Training for High Fidelity Natural Image Synthesis | TF PyTorch | ICLR |
Self-supervised adversarial training and high-resolution image synthesis with style incorporation (part 5)
Link to the article: Self-supervised adversarial training and high-resolution image synthesis with style incorporation part 5
First Author | Published | Title | Code | Conference/Journal |
---|---|---|---|---|
Ting Chen | 27 Nov 2018 | Self-Supervised GANs via Auxiliary Rotation Loss | PyTorch | CVPR |
Tero Karras | 12 Dec 2018 | A Style-Based Generator Architecture for Generative Adversarial Networks | TF PyTorch | CVPR |
Link to the article: Semantic image synthesis and learning a generative model from a single image part 6
First Author | Published | Title | Code | Conference/Journal |
---|---|---|---|---|
Taesung Park | 18 Mar 2019 | Semantic Image Synthesis with Spatially-Adaptive Normalization | PyTorch | CVPR |
Tamar Rott Shaham | 2 May 2019 | SinGAN: Learning a Generative Model from a Single Natural Image | PyTorch | ICCV |
- Play with Generative Adversarial Networks (GANs) in your browser!
- The GAN Zoo: A list of all named GANs
- Facebook AI GAN repo: A mix of GAN implementations including progressive growing
- PyTorch multiple implementations of Generative Adversarial Networks
- Another PyTorch GAN library that reproduces research results for popular GANs (CVPR 2020 Workshop)
- Keras implementations of Generative Adversarial Networks.
- Open Questions about Generative Adversarial Networks, Augustus Odena, Distill 2019
- Generative Adversarial Networks in Computer Vision: A Survey and Taxonomy
- From GAN to WGAN, Lil'Log blog 2017
@article{adaloglou2020gans,
title = "GANs in computer vision",
author = "Adaloglou, Nikolas and Karagiannakos, Sergios ",
journal = "https://theaisummer.com/",
year = "2020",
url = "https://theaisummer.com/gan-computer-vision/"
}