Home

Awesome

PUGAN_TIP2023

Runmin Cong, Wenyu Yang, Wei Zhang, Chongyi Li, Chun-Le Guo, Qingming Huang, and Sam Kwong, PUGAN: Physical model-guided underwater image enhancement using GAN with dual-discriminators, IEEE Transactions on Image Processing, vol. 32, pp. 4472-4485, 2023.

Network

Our overall framework:

image

Par-subnet:

image

TSIE-subnet:

image

Requirement:

Pleasure configure the environment according to the given version:

We also provide ".yaml" files for conda environment configuration, you can download it from [Link], code: mvpP, then use conda env create -f requirement.yaml to create a required environment.

Data Preprocessing

Please follow the tips to download the processed datasets and pre-trained model:

  1. Download training data from [Link], code: mvpP.
  2. Download testing data from [Link], code: mvpP.
├── utils
    ├── data_utils.py
├── Par
    ├── model
    ├── model.py
├── nets
    ├── pixpix.py
    ├── fusion.py
    ├── commons.py
├── checkpoints
├── test.py
├── train.py

Training and Testing

Training command : Please unzip the training data set to data\input_train and unzip the corresponding reference of training data set to data\gt_train.

We provide "train.yaml" files for training a new model from scratch or from a existing model.

python train.py

You can also train on a UFO or EVUP dataset by modifying train.yaml. We provide download connections for these datasets : UFO: [link], EUVP:[link]

Testing command : Please unzip the testing data set to tests.

We provide "test.yaml" files for testing.

The trained model can be download here: [[Link]n(https://pan.baidu.com/s/1y0_kHl1NRjrKc36LEX9wFQ?pwd=mvpP)], code: .mvpP

python test.py

Evaluation

We implement two metrics: PSNR, MSE.

python evaluations/measure_ssim_psnr.py

Results

  1. Qualitative results: we provide the saliency maps, you can download them from [Link], code: mvpP.
  2. Quantitative results:

image

Bibtex

   @article{crm/tip23/PUGAN,
           author={Cong, Runmin and Yang, Wenyu and Zhang, Wei and Li, chongyi and Guo, Chun-Le and Huang, Qingming and Kwong, Sam },
           journal={IEEE Trans. Image Process. },
           title= {{PUGAN}: Physical model-guided underwater image enhancement using {GAN} with dual-discriminators},
           volume={32},
           pages={4472-4485},
           year={2023}
           }

Contact Us

If you have any questions, please contact Runmin Cong at rmcong@sdu.edu.cn or Wenyu Yang at wyuyang@bjtu.edu.cn.