[CVPR 2023] Learning Semantic-Aware Knowledge Guidance for Low-Light Image Enhancement (Paper)

Yuhui Wu, Chen Pan, Guoqing Wang*, Yang Yang, Jiwei Wei, Chongyi Li, Heng Tao Shen (*Corresponding Author)

University of Electronic Science and Technology of China (UESTC)

Introduction

This repository is the official implementation of the paper, "Learning Semantic-Aware Knowledge Guidance for Low-Light Image Enhancement", where more implementation details are presented.

Motivation and Superiority

Overall

Dataset

You can refer to the following links to download the datasets:

LOL: Chen Wei, Wenjing Wang, Wenhan Yang, and Jiaying Liu. "Deep Retinex Decomposition for Low-Light Enhancement", BMVC, 2018. [Baiduyun (extracted code: sdd0)] [Google Drive]
LOL-v2 (the extension work): Wenhan Yang, Haofeng Huang, Wenjing Wang, Shiqi Wang, and Jiaying Liu. "Sparse Gradient Regularized Deep Retinex Network for Robust Low-Light Image Enhancement", TIP, 2021. [Baiduyun (extracted code: l9xm)] [Google Drive]

Results

The evauluation results on LOL are as follows:

Method	PSNR	SSIM	LPIPS	Method	PSNR	SSIM	LPIPS
RetinexNet	16.770	0.462	0.474	HWMNet	24.240	0.852	0.114
RetinexNet-SKF	20.418	0.711	0.216	HWMNet-SKF	25.086	0.860	0.108
KinD	20.870	0.799	0.207	SNR-LLIE-Net	24.608	0.840	0.151
KinD-SKF	21.913	0.835	0.143	SNR-LLIE-Net-SKF	25.031	0.855	0.113
DRBN	19.860	0.834	0.155	LLFlow-S	24.060	0.860	0.136
DRBN-SKF	22.837	0.841	0.138	LLFlow-S-SKF	25.942	0.865	0.125
KinD++	18.970	0.804	0.175	LLFlow-L	24.999	0.870	0.117
KinD++-SKF	20.363	0.805	0.201	LLFlow-L-SKF	26.798	0.879	0.105

Test and Train

More details can be found in the subfolder of each baseline.

Citation

If you find our work useful for your research, please cite our paper

@inproceedings{wu2023skf,
  title={Learning Semantic-Aware Knowledge Guidance for Low-Light Image Enhancement},
  author={Yuhui, Wu and Chen, Pan and Guoqing, Wang and Yang, Yang and Jiwei, Wei and Chongyi, Li and Heng Tao Shen},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={},
  year={2023}
}

Contact

If you have any question, please feel free to contact us via wuyuhui132@gmail.com or panchen0103@gmail.com.