https://github.com/hunterlarco/ghilbi-gan
A style-transfer project using CycleGAN to render photos in the style of Studio Ghibli animations.
https://github.com/hunterlarco/ghilbi-gan
anime gans ghibli machine-learning neural-network python tensorflow
Last synced: 4 months ago
JSON representation
A style-transfer project using CycleGAN to render photos in the style of Studio Ghibli animations.
- Host: GitHub
- URL: https://github.com/hunterlarco/ghilbi-gan
- Owner: HunterLarco
- License: mit
- Created: 2019-02-18T23:58:37.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2019-02-20T01:22:37.000Z (over 7 years ago)
- Last Synced: 2025-04-30T14:32:30.951Z (about 1 year ago)
- Topics: anime, gans, ghibli, machine-learning, neural-network, python, tensorflow
- Language: Python
- Homepage:
- Size: 12.7 KB
- Stars: 10
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# GhibliGAN
A style-transfer project using [CycleGAN] to render photos in the style of
Studio Ghibli animations.
> Note that this project is built as a learning project for myself and that if
> you're looking for a tried-and-true implementation of CycleGAN you should
> refer to their [original PyTorch source][CycleGANSource] or look at one of the
> several [Tensorflow adaptations][CycleGANTensorflow].
## Build Setup
```bash
# install package manager
pip install pipenv
# install dependencies
pipenv install
```
## Code Quality
```bash
# get linter aggregate statistics
pipenv run pycodestyle --statistics -qq --config pycodestyle.cfg src
# lint a specific file(s)
pipenv run pycodestyle --first --show-source --config pycodestyle.cfg
```
## Model Architecture
The base for this model is [CycleGAN]—the generator network follows
[Johnson et al.] while the discriminator follows [PatchGAN].
[CycleGAN]: https://arxiv.org/abs/1703.10593
[CycleGANSource]: https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix
[CycleGANTensorflow]: https://github.com/search?q=CycleGan+tensorflow
[Johnson et al.]: https://arxiv.org/abs/1603.08155
[PatchGAN]: https://arxiv.org/abs/1611.07004