GuidedHybSensUIR

Underwater Image Restoration Through a Prior Guided Hybrid Sense Approach and Extensive Benchmark Analysis

<div> <span class="author-block"> Xiaojiao Guo^{👨‍💻‍} </span>, <span class="author-block"> <a href='https://cxh.netlify.app/'>Xuhang Chen</a>^{👨‍💻‍} </span>, <span class="author-block"> Shuqiang Wang^📮 </span>, <span class="author-block"> <a href='https://cmpun.github.io/'>Chi-Man Pun</a>^📮 </span> ( 👨‍💻‍ Equal contributions, 📮 Corresponding author) </div>

<b>University of Macau, SIAT CAS, Huizhou Univeristy, Baoshan Univeristy</b>

In <b>IEEE Transactions on Circuits and Systems for Video Technology</b>

🔮 Benchmark Dataset

Kaggle

If you need visualization results, you may contact Dr.Guo via yc27441@um.edu.mo

⚙️ Usage

Training

You may download the dataset first, and then specify TRAIN_DIR, VAL_DIR and SAVE_DIR in the section TRAINING in config.yml.

For single GPU training:

python train.py

For multiple GPUs training:

accelerate config
accelerate launch train.py

If you have difficulties with the usage of accelerate, please refer to <a href="https://github.com/huggingface/accelerate">Accelerate</a>.

Inference

Please first specify TRAIN_DIR, VAL_DIR and SAVE_DIR in section TESTING in config.yml.

python infer.py

Citation

@ARTICLE{10824878,
  author={Guo, Xiaojiao and Chen, Xuhang and Wang, Shuqiang and Pun, Chi-Man},
  journal={IEEE Transactions on Circuits and Systems for Video Technology}, 
  title={Underwater Image Restoration Through a Prior Guided Hybrid Sense Approach and Extensive Benchmark Analysis}, 
  year={2025},
  volume={},
  number={},
  pages={1-1},
  keywords={Image restoration;Image color analysis;Benchmark testing;Deep learning;Green products;Distortion;Transformers;Adaptation models;Training;Lighting;Underwater image restoration;image enhancement;prior guided attention;efficient Transformer;multi-scales hybridization},
  doi={10.1109/TCSVT.2025.3525593}
}