ESRGAN-PyTorch

Overview

This repository contains an op-for-op PyTorch reimplementation of ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks.

Table of contents

• ESRGAN-PyTorch
  • Overview
  • Table of contents
  • Download weights
  • Download datasets
  • How Test and Train
    • Test ESRGAN_x4
    • Train RRDBNet_x4
    • Resume train RRDBNet_x4
    • Train ESRGAN_x4
    • Resume train ESRGAN_x4
  • Result
  • Contributing
  • Credit
    • ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks

Download weights

• Google Driver
• Baidu Driver

Download datasets

Contains DIV2K, DIV8K, Flickr2K, OST, T91, Set5, Set14, BSDS100 and BSDS200, etc.

• Google Driver
• Baidu Driver

Please refer to README.md in the data directory for the method of making a dataset.

How Test and Train

Both training and testing only need to modify yaml file.

Test ESRGAN_x4

python3 test.py --config_path ./configs/test/ESRGAN_x4-DFO2K-Set5.yaml

Train RRDBNet_x4

python3 train_net.py --config_path ./configs/train/RRDBNet_x4-DFO2K.yaml

Resume train RRDBNet_x4

Modify the ./configs/train/RRDBNet_X4.yaml file.

• line 34: RESUMED_G_MODEL change to ./samples/RRDBNet_X4-DIV2K/g_epoch_xxx.pth.tar.

python3 train_net.py --config_path ./configs/train/RRDBNet_x4-DFO2K.yaml

Train ESRGAN_x4

Modify the ./configs/train/ESRGAN_X4.yaml file.

• line 39: PRETRAINED_G_MODEL change to ./results/EDSRGAN_x4-DIV2K/g_last.pth.tar.

python3 train_gan.py --config_path ./configs/train/ESRGAN_x4-DFO2K.yaml

Resume train ESRGAN_x4

Modify the ./configs/train/ESRGAN_X4.yaml file.

• line 39: PRETRAINED_G_MODEL change to ./results/RRDBNet_x4-DIV2K/g_last.pth.tar.
• line 41: RESUMED_G_MODEL change to ./samples/EDSRGAN_x4-DIV2K/g_epoch_xxx.pth.tar.
• line 42: RESUMED_D_MODEL change to ./samples/EDSRGAN_x4-DIV2K/d_epoch_xxx.pth.tar.

python3 train_gan.py --config_path ./configs/train/ESRGAN_x4-DFO2K.yaml

Result

Source of original paper results: https://arxiv.org/pdf/1809.00219v2.pdf

In the following table, the value in () indicates the result of the project, and - indicates no test.

Method	Scale	Set5 (PSNR/SSIM)	Set14 (PSNR/SSIM)	BSD100 (PSNR/SSIM)	Urban100 (PSNR/SSIM)	Manga109 (PSNR/SSIM)
RRDB	4	32.73(32.71)/0.9011(0.9018)	28.99(28.96)/0.7917(0.7917)	27.85(27.85)/0.7455(0.7473)	27.03(27.03)/0.8153(0.8156)	31.66(31.60)/0.9196(0.9195)
ESRGAN	4	-(30.44)/-(0.8525)	-(26.28)/-(0.6994)	-(25.33)/-(0.6534)	-(24.36)/-(0.7341)	-(29.42)/-(0.8597)

# Download `ESRGAN_x4-DFO2K-25393df7.pth.tar` weights to `./results/pretrained_models`
# More detail see `README.md<Download weights>`
python3 ./inference.py

Input:

Output:

Build `rrdbnet_x4` model successfully.
Load `rrdbnet_x4` model weights `/ESRGAN-PyTorch/results/pretrained_models/ESRGAN_x4-DFO2K.pth.tar` successfully.
SR image save to `./figure/ESRGAN_x4_baboon.png`

Contributing

If you find a bug, create a GitHub issue, or even better, submit a pull request. Similarly, if you have questions, simply post them as GitHub issues.

I look forward to seeing what the community does with these models!

Credit

ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks

Xintao Wang, Ke Yu, Shixiang Wu, Jinjin Gu, Yihao Liu, Chao Dong, Chen Change Loy, Yu Qiao, Xiaoou Tang

Abstract
The Super-Resolution Generative Adversarial Network (SRGAN) is a seminal work that is capable of generating realistic textures during single image super-resolution. However, the hallucinated details are often accompanied with unpleasant artifacts. To further enhance the visual quality, we thoroughly study three key components of SRGAN - network architecture, adversarial loss and perceptual loss, and improve each of them to derive an Enhanced SRGAN (ESRGAN). In particular, we introduce the Residual-in-Residual Dense Block (RRDB) without batch normalization as the basic network building unit. Moreover, we borrow the idea from relativistic GAN to let the discriminator predict relative realness instead of the absolute value. Finally, we improve the perceptual loss by using the features before activation, which could provide stronger supervision for brightness consistency and texture recovery. Benefiting from these improvements, the proposed ESRGAN achieves consistently better visual quality with more realistic and natural textures than SRGAN and won the first place in the PIRM2018-SR Challenge. The code is available at this https URL.

[Paper] [Author's implements(PyTorch)]

@misc{wang2018esrgan,
    title={ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks},
    author={Xintao Wang and Ke Yu and Shixiang Wu and Jinjin Gu and Yihao Liu and Chao Dong and Chen Change Loy and Yu Qiao and Xiaoou Tang},
    year={2018},
    eprint={1809.00219},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}