https://github.com/raphaelsenn/pytorch-gans
PyTorch implementations of various GAN papers/architectures like GAN (Goodfellow at at.), DCGAN (Radford et al.), etc.
https://github.com/raphaelsenn/pytorch-gans
Last synced: 11 months ago
JSON representation
PyTorch implementations of various GAN papers/architectures like GAN (Goodfellow at at.), DCGAN (Radford et al.), etc.
- Host: GitHub
- URL: https://github.com/raphaelsenn/pytorch-gans
- Owner: raphaelsenn
- License: mit
- Created: 2025-07-14T18:34:50.000Z (11 months ago)
- Default Branch: main
- Last Pushed: 2025-07-27T17:38:32.000Z (11 months ago)
- Last Synced: 2025-07-27T19:29:39.284Z (11 months ago)
- Language: Python
- Size: 142 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# pytorch-GANs
PyTorch implementations of various GAN papers/architectures like GAN (Goodfellow at at.), DCGAN (Radford et al.), etc.
## Implemented Papers
* [Generative Adversarial Networks (Goodfellow et al., 2014)](https://arxiv.org/abs/1406.2661)
* [Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks (Radford et al., 2016)](https://arxiv.org/abs/1511.06434)
* [Conditional Generative Adversarial Nets (Mirza et al., 2014)](https://arxiv.org/abs/1411.1784)
## Generative Adversarial Networks (Goodfellow et al., 2014)
This code tries to reproduce the 2014 Goodfellow et al. paper [Generative Adversarial Networks](https://arxiv.org/abs/1406.2661).
Original code and hyperparameters from the paper can be found [here](https://github.com/goodfeli/adversarial).
### MNIST (fully connected model)
#### Model Architecture:
Generator:
* Input layer: 100 units (Uniform[-1, 1] noise)
* Dense layer: 1200 units, ReLU
* Dense layer: 1200 units, ReLU
* Output layer: 784 units, Sigmoid
Discriminator:
* Input layer: 784 units (flattend 28x28 gray-scale image)
* Maxout layer: 240 units, 5 pieces
* Maxout layer: 240 units, 5 pieces
* Output layer: 1 unit, Sigmoid
| Epoch 001 | Epoch 100 |Loss |
| --------- | ---------- |-----------|
|  | | |
### CelebFaces (fully connected model)
#### Model Architecture:
Generator:
* Input layer: 100 units (Uniform[-1, 1] noise)
* Dense layer: 8000 units, ReLU
* Output layer: 2304 units, Sigmoid
Discriminator:
* Input layer: 2304 units (flattend 48x48 grayscale image)
* Maxout layer: 1200 units, 5 pieces
* Maxout layer: 1200 units, 5 pieces
* Output layer: 1 unit, Sigmoid
| Epoch 001 | Epoch 250 |Loss |
| --------- | ---------- |-----------|
|  | | |
## Deep Convolutional Generative Adversarial Networks (Radford et al. 2016)
This code tries to reproduce the 2016 Radford et al. paper [Deep Convolutional Generative Adversarial Networks](https://arxiv.org/abs/1511.06434).
*Figure 1: The architecture of the DCGAN generator used for CelebFaces modeling; Taken from Radford et al., 2016.*
#### CelebFaces (deep convolutional model)
| Epoch 01 | Epoch 50 | Loss |
| --------- | ---------- | ------- |
|  |  | |
##### About the CelebFaces dataset
The dataset was taken from kaggle.
Link to the dataset can be found [here](https://www.kaggle.com/datasets/jessicali9530/celeba-dataset).
Reproducibility:
Download the [CelebA dataset](https://mmlab.ie.cuhk.edu.hk/projects/CelebA.html) and place the images and attribute CSV in a folder called `celeba`.
```python
dataset = CelebA(csv_file="list_attr_celeba.csv", root_dir="img_align_celeba", transform=...)
```
## Conditional Generative Adversarial Nets
This code tries to reproduce the 2014 Mirza et al. paper [Conditional Generative Adversarial Nets](https://arxiv.org/abs/1411.1784).
*Architecture of an conditional generative adversarial net; Taken from Mirza et al., 2014.*
#### MNIST (fully connected model)
| Epoch 001 | Epoch 100 | Loss |
| --------- | ---------- | ------- |
|  |  | |
## Citations
```bibtex
@misc{goodfellow2014generativeadversarialnetworks,
title={Generative Adversarial Networks},
author={Ian J. Goodfellow and Jean Pouget-Abadie and Mehdi Mirza and Bing Xu and David Warde-Farley and Sherjil Ozair and Aaron Courville and Yoshua Bengio},
year={2014},
eprint={1406.2661},
archivePrefix={arXiv},
primaryClass={stat.ML},
url={https://arxiv.org/abs/1406.2661},
}
```
```bibtex
@misc{radford2016unsupervisedrepresentationlearningdeep,
title={Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks},
author={Alec Radford and Luke Metz and Soumith Chintala},
year={2016},
eprint={1511.06434},
archivePrefix={arXiv},
primaryClass={cs.LG},
url={https://arxiv.org/abs/1511.06434},
}
```
```bibtex
@misc{mirza2014conditionalgenerativeadversarialnets,
title={Conditional Generative Adversarial Nets},
author={Mehdi Mirza and Simon Osindero},
year={2014},
eprint={1411.1784},
archivePrefix={arXiv},
primaryClass={cs.LG},
url={https://arxiv.org/abs/1411.1784},
}
```
```bibtex
@misc{liu2015celeba,
title = {CelebA: Large-scale CelebFaces Attributes Dataset},
author = {Z. Liu and P. Luo and X. Wang and X. Tang},
year = {2015},
howpublished = {\url{https://www.kaggle.com/datasets/jessicali9530/celeba-dataset}},
note = {Original dataset by MMLAB, The Chinese University of Hong Kong},
institution = {Kaggle},
keywords = {face recognition, facial attributes, dataset}
}
```