The code of "WV-LUT: Wide Vision Lookup Tables for Real-Time Low-Light Image Enhancement".
python=3.8
requirements.txt
data/
├── LOL_v1
│ ├── high
│ └── low
├── ...
└── LowLightBenchmark
├── ...
└── LOL_v1_val
├── high
└── low
Download the required data here, the dataset directory structure is as shown above
# train net
sh script/train.sh
# build LUT
sh script/transferLUT.sh
# finetune LUT
sh script/finetuneLUT.sh
# evaluation
sh script/eval.sh
@ARTICLE{wvlut,
author={Li, Canlin and Su, Haowen and Tan, Xin and Zhang, Xiangfei and Ma, Lizhuang},
journal={IEEE Transactions on Multimedia},
title={WV-LUT: Wide Vision Lookup Tables for Real-Time Low-Light Image Enhancement},
year={2025},
doi={10.1109/TMM.2025.3535342}
}
Feel free to contact us with any questions(email).
This code is build upon MuLUT and IAT. Thank them for their excellent works.