https://github.com/woctezuma/steam-stylegan2-ada-pytorch

Train a StyleGAN2-ADA (PyTorch) model on Colab to generate Steam banners.
https://github.com/woctezuma/steam-stylegan2-ada-pytorch

colab colab-notebook colaboratory gan generative-adversarial-network google-colab google-colab-notebook google-colaboratory pytorch steam steam-data steam-game steam-games steam-gan steam-store stylegan stylegan-model stylegan2 stylegan2-ada stylegan2-ada-pytorch

Last synced: 20 days ago
JSON representation

Train a StyleGAN2-ADA (PyTorch) model on Colab to generate Steam banners.

• Host: GitHub
• URL: https://github.com/woctezuma/steam-stylegan2-ada-pytorch
• Owner: woctezuma
• License: mit
• Created: 2021-02-01T13:15:26.000Z (almost 4 years ago)
• Default Branch: main
• Last Pushed: 2024-01-10T12:34:25.000Z (12 months ago)
• Last Synced: 2024-01-10T14:42:56.402Z (12 months ago)
• Topics: colab, colab-notebook, colaboratory, gan, generative-adversarial-network, google-colab, google-colab-notebook, google-colaboratory, pytorch, steam, steam-data, steam-game, steam-games, steam-gan, steam-store, stylegan, stylegan-model, stylegan2, stylegan2-ada, stylegan2-ada-pytorch
• Language: Jupyter Notebook
• Homepage: https://colab.research.google.com/
• Size: 494 KB
• Stars: 8
• Watchers: 3
• Forks: 3
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Steam StyleGAN2-ADA (PyTorch)

The goal of this repository is to capture the distribution of Steam banners with a StyleGAN2-ADA model (PyTorch).

![Target A][target-A]![Projection A][projection-A]

![Target B][target-B]![Projection B][projection-B]



Progression videos of the projection of Steam banners ([*Dog Trainer*][store-link-A] and [*My UnReal pet*][store-link-B]) with a network pre-trained by Nvidia on the `LSUN DOG` dataset.



## Data

The [`Steam-OneFace`][steam-oneface] dataset,

obtained with the `retinaface` face detection module:

-   consists of 2.5k Steam banners (256x256 resolution) which should all feature exactly one face,

-   can be downloaded (74 MB) with the following commands:

```bash

!gdown --id 1-0Nk7H6Cn3Nt60EdHG_NWSA8ohi2oBqr

!tar xf steam-oneface-lr_with_retinaface.tar.gz

```

## Usage

-   Run [`training.ipynb`][colab-notebook-training] to train a model from scratch,

[![Open In Colab][colab-badge]][colab-notebook-training]

-   Run [`image_sampling.ipynb`][colab-notebook-sampling] to generate images with a trained model,

[![Open In Colab][colab-badge]][colab-notebook-sampling]

-   To automatically resume training from the latest checkpoint, use [my fork][stylegan2-ada-pytorch-fork] of StyleGAN2-ADA (PyTorch).

## References

-   [Karras, Tero, et al. *Training generative adversarial networks with limited data*. NeurIPS 2020][stylegan2-ada-paper],

-   [Official implementation (PyTorch)][stylegan2-ada-pytorch-repository],

-   Application to Steam banners using the [TensorFlow implementation][stylegan2-ada-applied-to-steam-banners].

[target-A]: 

[projection-A]: 

[store-link-A]: 

[target-B]: 

[projection-B]: 

[store-link-B]: 

[steam-oneface]: 

[colab-notebook-training]: 

[colab-notebook-sampling]: 

[colab-badge]: 

[stylegan2-ada-pytorch-fork]: 

[stylegan2-ada-paper]: 

[stylegan2-ada-pytorch-repository]: 

[stylegan2-ada-applied-to-steam-banners]: