mhmdjouni/mog-dun

PyTorch code for JSTSP2021 paper "Accurate and Lightweight Image Super-Resolution with Model-Guided Deep Unfolding Network""

Accurate and Lightweight Image Super-Resolution with Model-Guided Deep Unfolding Network [IEEE] [homepage]

This repository is Pytorch code for our proposed MoG-DUN.

The code is built on RCAN and tested on Ubuntu 16.04 environment (Python 3.5/3.6/3.7, PyTorch 1.0.0/1.0.1, 9.0/10.0) with 2080Ti/1080Ti GPUs.

If you find our work useful in your research or publications, please consider citing:

@ARTICLE{9257009,
  author={Q. {Ning} and W. {Dong} and G. {Shi} and L. {Li} and X. {Li}},
  journal={IEEE Journal of Selected Topics in Signal Processing}, 
  title={Accurate and Lightweight Image Super-Resolution with Model-Guided Deep Unfolding Network}, 
  year={2020},
  volume={},
  number={},
  pages={1-1},
  doi={10.1109/JSTSP.2020.3037516}}

Contents

1. Requirements
2. Test
3. Acknowledgements

Requirements

• Python 3
• skimage
• imageio
• Pytorch (Pytorch version 1.0.1 is recommended)
• tqdm
• cv2 (pip install opencv-python)

Test

Quick start

Test on standard SR benchmark

1. If you have cloned this repository, the pre-trained models can be found in experiment fold and test dataset Set5 can be found in data fold.

2. Then, run command:

   cd code_AG
   sh test.sh
   

3. Finally, PSNR values are shown on your screen, you can find the reconstruction images in ../experiment/xx/results/

Acknowledgements

• This code is built on RCAN (PyTorch). We thank the authors for sharing their codes.