Skip to content

Latest commit

 

History

History
111 lines (79 loc) · 6.09 KB

README.md

File metadata and controls

111 lines (79 loc) · 6.09 KB
Raw

CiaoSR: Continuous Implicit Attention-in-Attention Network for Arbitrary-Scale Image Super-Resolution (CVPR 2023)

Jiezhang Cao, Qin Wang, Yongqin Xian, Yawei Li, Bingbing Ni, Zhiming Pi, Kai Zhang, Yulun Zhang, Radu Timofte, Luc Van Gool

Computer Vision Lab, ETH Zurich.

arxiv | supplementary | pretrained models | visual results

arXiv GitHub Stars download visitors

This repository is the official PyTorch implementation of "CiaoSR: Continuous Implicit Attention-in-Attention Network for Arbitrary-Scale Image Super-Resolution" (arxiv, supp, pretrained models, visual results). CiaoSR achieves state-of-the-art performance in arbitrary-scale image super-resolution.

Learning continuous image representations is recently gaining popularity for image super-resolution (SR) because of its ability to reconstruct high-resolution images with arbitrary scales from low-resolution inputs. Existing methods mostly ensemble nearby features to predict the new pixel at any queried coordinate in the SR image. Such a local ensemble suffers from some limitations: i) it has no learnable parameters and it neglects the similarity of the visual features; ii) it has a limited receptive field and cannot ensemble relevant features in a large field which are important in an image. To address these issues, this paper proposes a continuous implicit attention-in-attention network, called CiaoSR. We explicitly design an implicit attention network to learn the ensemble weights for the nearby local features. Furthermore, we embed a scale-aware attention in this implicit attention network to exploit additional non-local information. Extensive experiments on benchmark datasets demonstrate CiaoSR significantly outperforms the existing single image SR methods with the same backbone. In addition, CiaoSR also achieves the state-of-the-art performance on the arbitrary-scale SR task. The effectiveness of the method is also demonstrated on the real-world SR setting. More importantly, CiaoSR can be flexibly integrated into any backbone to improve the SR performance.

Contents

  1. Requirements
  2. Quick Testing
  3. Training
  4. Results
  5. Citation
  6. License and Acknowledgement

TODO

  • Add pretrained model
  • Add results of test set
  • Add real-world arbitrary-scale SR

Requirements

  • Platforms: Ubuntu 18.04, cuda-11.1
  • Python 3.8, PyTorch >= 1.9.1
  • mmedit 0.11.0
# install mmcv
pip install --no-cache-dir --ignore-installed mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu110/torch1.9.1/index.html 

# download mmediting from https://openi.pcl.ac.cn/open-mmlab/mmediting.git
cd mmediting
python setup.py develop

Quick Testing

Following commands will download pretrained models and test datasets. If out-of-memory, try to reduce tile at the expense of slightly decreased performance.

# download code
git clone https://github.com/caojiezhang/CiaoSR
cd CiaoSR


PYTHONPATH=/bin/..:tools/..: python tools/test.py configs/001_localimplicitsr_rdn_div2k_g1_c64b16_1000k_unfold_lec_mulwkv_res_nonlocal.py pretrain_model/rdn-ciaosr.pth

All visual results of CiaoSR can be downloaded here.

Training

PYTHONPATH=/bin/..:tools/..: ./tools/dist_train.sh configs/001_localimplicitsr_rdn_div2k_g1_c64b16_1000k_unfold_lec_mulwkv_res_nonlocal.py 8

Results

We achieved state-of-the-art performance on arbitrary-scale iamge super-resolution. Detailed results can be found in the paper.

Citation

@inproceedings{cao2023ciaosr,
  title={CiaoSR: Continuous Implicit Attention-in-Attention Network for Arbitrary-Scale Image Super-Resolution},
  author={Cao, Jiezhang and Wang, Qin and Xian, Yongqin and Li, Yawei and Ni, Bingbing and Pi, Zhiming and Zhang, Kai and Zhang, Yulun and Timofte, Radu and Van Gool, Luc},
  booktitle={IEEE Conference on Computer Vision and Pattern Recognition},
  year={2023}
}

License and Acknowledgement

This project is released under the CC-BY-NC license. We refer to codes from KAIR, BasicSR, and mmediting. Thanks for their awesome works. The majority of CiaoSR is licensed under CC-BY-NC, however portions of the project are available under separate license terms: KAIR is licensed under the MIT License, BasicSR, and mmediting are licensed under the Apache 2.0 license.