[IMAVIS'25] UNIR-Net: A Novel Approach for Restoring Underwater Images with Non-Uniform Illumination Using Synthetic Data

🎯 1. Overview

This repository contains the source code and supplementary materials for the paper titled UNIR-Net: A Novel Approach for Restoring Underwater Images with Non-Uniform Illumination Using Synthetic Data. This research focuses on the visual enhancement of underwater images with non-uniform illumination. The paper has been accepted for publication in Image and Vision Computing (IMAVIS).

🛠️ 2. Requirements

1. opencv-python == 4.9.0.80
2. scikit-image == 0.22.0
3. numpy == 1.24.3
4. torch == 2.3.0+cu118
5. Pillow == 10.2.0
6. tqdm == 4.65.0
7. natsort == 8.4.0
8. torchvision == 0.18.0+cu118

🧪 3. Inference

To test the model, follow these steps:

1. Place your images to be enhanced in the ./1_Input directory.

2. Run the code with the following command:

   python inference.py
   

3. The enhanced images will be saved in the ./2_Output directory.

📄 Citation

If this work contributes to your research, we would appreciate it if you could cite our paper:

@article{perez2025unir,
  title={UNIR-Net: A novel approach for restoring underwater images with non-uniform illumination using synthetic data},
  author={Perez-Zarate, Ezequiel and Liu, Chunxiao and Ramos-Soto, Oscar and Oliva, Diego and Perez-Cisneros, Marco},
  journal={Image and Vision Computing},
  pages={105734},
  year={2025},
  publisher={Elsevier}
}