/ Machine Learning

Generative Adversarial Networks with Ranker for Image Super-Resolution

Generative Adversarial Networks with Ranker for Image Super-Resolution

RankSRGAN

ICCV 2019 (oral) RankSRGAN: Generative Adversarial Networks with Ranker for Image Super-Resolution.

Dependencies

Codes

This version is under testing. We will provide more details of RankSRGAN in near future.

How to Test

  1. Clone this github repo.
git clone https://github.com/WenlongZhang0724/RankSRGAN.git
cd RankSRGAN
  1. Place your own low-resolution images in ./LR folder.
  2. Download pretrained models from Google Drive or Baidu Drive. Place the models in ./experiments/pretrained_models/. We provide three Ranker models and three RankSRGAN models (see model list).
  3. Run test. We provide RankSRGAN (NIQE, Ma, PI) model and you can config in the test.py.
python test.py -opt options/test/test_ranksrgan.json
  1. The results are in ./results folder.

How to Train

Train Ranker

  1. Download DIV2K and Flickr2K from Google Drive or Baidu Drive
  2. Generate rank dataset ./datascripts/generate_rankdataset/
  3. Run command:
python train.py -opt options/train/Ranker.json

Train RankSRGAN

We use a PSNR-oriented pretrained SR model to initialize the parameters for better quality.

  1. Prepare datasets, usually the DIV2K dataset.
  2. Prerapre the PSNR-oriented pretrained model. You can use the SRResNet_bicx4_in3nf64nb16.pth as the pretrained model that can be downloaded from Google Drive or Baidu Drive.
  3. Modify the configuration file options/train/RankSRGAN_NIQE.json
  4. Run command:
python train.py -opt options/train/RankSRGAN_NIQE.json

or

python train_PI.py -opt options/train/RankSRGAN_NIQE.json

Using the train.py can output the convergence curves with NIQE and PSNR; Using the train_PI.py can output the convergence curves with NIQE, Ma, PI and PSNR.

GitHub