# GANetic
A collection of GANs implemented in PyTorch.
## Table of Contents
<!-- - [Installation](#installation) -->
- [Usage](#usage)
- [DCGAN](#dcgan)
- [SRGAN](#srgan)
- [Citations](#citations)
## Usage
### DCGAN
```python
import torch
from ganetic.dcgan import Discriminator, Generator
netG = Generator(
nz=100, # length of latent vector
nc=3, # number of channels in the training images.
ngf=64, # size of feature maps in generator
)
netD = Discriminator(
nc=3, # number of channels in the training images.
ndf=64, # size of feature maps in discriminator
)
noise = torch.randn(1, 100, 1, 1)
fake_img = netG(noise)
prediction = netD(fake_img)
```
### SRGAN
```python
import torch
from ganetic.srgan import Generator, Discriminator
img = torch.randn(1, 3, 64, 64)
gen = Generator(
scale_factor=4, # scale factor for super resolution
nci=3, # number of channels in input image
nco=3, # number of channels in output image
ngf=64, # number of filters in the generator
no_of_residual_blocks=5
)
disc = Discriminator(
input_shape=(3, 256, 256),
ndf=64, # number of filters in the discriminator
negative_slope=0.2, # negative slope of leaky relu
)
HR_img = gen(img)
pred = disc(HR_img)
```
## Citations
```bibtex
@article{radford2015unsupervised,
title={Unsupervised representation learning with deep convolutional generative adversarial networks},
author={Radford, Alec and Metz, Luke and Chintala, Soumith},
journal={arXiv preprint arXiv:1511.06434},
year={2015}
}
```
```bibtex
@inproceedings{ledig2017photo,
title={Photo-realistic single image super-resolution using a generative adversarial network},
author={Ledig, Christian and Theis, Lucas and Husz{\'a}r, Ferenc and Caballero, Jose and Cunningham, Andrew and Acosta, Alejandro and Aitken, Andrew and Tejani, Alykhan and Totz, Johannes and Wang, Zehan and others},
booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
pages={4681--4690},
year={2017}
}
```
Raw data
{
"_id": null,
"home_page": "https://github.com/kingjuno/GANetic",
"name": "GANetic",
"maintainer": "",
"docs_url": null,
"requires_python": "",
"maintainer_email": "",
"keywords": "GAN,Generative Adversarial Networks,Deep Learning,PyTorch,GANetic",
"author": "Geo Jolly",
"author_email": "geojollyc@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/59/cc/9e9659d50b30e6b8ae43929fa6f93aad8d849a39e5470f9e66ee364d034c/GANetic-0.0.3.tar.gz",
"platform": null,
"description": "# GANetic\nA collection of GANs implemented in PyTorch.\n\n## Table of Contents\n<!-- - [Installation](#installation) -->\n- [Usage](#usage)\n - [DCGAN](#dcgan)\n - [SRGAN](#srgan)\n- [Citations](#citations)\n\n## Usage\n### DCGAN\n```python\nimport torch\n\nfrom ganetic.dcgan import Discriminator, Generator\n\nnetG = Generator(\n nz=100, # length of latent vector\n nc=3, # number of channels in the training images.\n ngf=64, # size of feature maps in generator\n)\nnetD = Discriminator(\n nc=3, # number of channels in the training images.\n ndf=64, # size of feature maps in discriminator\n)\n\nnoise = torch.randn(1, 100, 1, 1)\nfake_img = netG(noise)\nprediction = netD(fake_img)\n```\n\n### SRGAN\n```python\nimport torch\n\nfrom ganetic.srgan import Generator, Discriminator\n\nimg = torch.randn(1, 3, 64, 64)\ngen = Generator(\n scale_factor=4, # scale factor for super resolution\n nci=3, # number of channels in input image\n nco=3, # number of channels in output image\n ngf=64, # number of filters in the generator\n no_of_residual_blocks=5 \n)\ndisc = Discriminator(\n input_shape=(3, 256, 256),\n ndf=64, # number of filters in the discriminator\n negative_slope=0.2, # negative slope of leaky relu\n)\n\nHR_img = gen(img)\npred = disc(HR_img)\n```\n\n## Citations\n```bibtex\n@article{radford2015unsupervised,\n title={Unsupervised representation learning with deep convolutional generative adversarial networks},\n author={Radford, Alec and Metz, Luke and Chintala, Soumith},\n journal={arXiv preprint arXiv:1511.06434},\n year={2015}\n}\n```\n\n```bibtex\n@inproceedings{ledig2017photo,\n title={Photo-realistic single image super-resolution using a generative adversarial network},\n author={Ledig, Christian and Theis, Lucas and Husz{\\'a}r, Ferenc and Caballero, Jose and Cunningham, Andrew and Acosta, Alejandro and Aitken, Andrew and Tejani, Alykhan and Totz, Johannes and Wang, Zehan and others},\n booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},\n pages={4681--4690},\n year={2017}\n}\n```\n\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "A Collection of GANs - PyTorch",
"version": "0.0.3",
"split_keywords": [
"gan",
"generative adversarial networks",
"deep learning",
"pytorch",
"ganetic"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "59cc9e9659d50b30e6b8ae43929fa6f93aad8d849a39e5470f9e66ee364d034c",
"md5": "6dd1c59305070bc61b876217dcb0e807",
"sha256": "52924e19d8cc6fea4d3da654ff4fd37a1384264306efb769ada2e57a14a84454"
},
"downloads": -1,
"filename": "GANetic-0.0.3.tar.gz",
"has_sig": false,
"md5_digest": "6dd1c59305070bc61b876217dcb0e807",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 4598,
"upload_time": "2023-02-08T15:30:49",
"upload_time_iso_8601": "2023-02-08T15:30:49.720582Z",
"url": "https://files.pythonhosted.org/packages/59/cc/9e9659d50b30e6b8ae43929fa6f93aad8d849a39e5470f9e66ee364d034c/GANetic-0.0.3.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-02-08 15:30:49",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "kingjuno",
"github_project": "GANetic",
"lcname": "ganetic"
}
JSON
