EENet: An effective and efficient network for single image dehazing (PR2025)

Download the Datasets

• SOTS [gdrive, Baidu]

Pre-trained models and visual results

ITS/OTS: Gdrive

CSD/Snow100K/SRRS: 百度网盘

If you require models or results for other tasks, please feel free to contact me via issue or email. Thank you.

Installation

The project is built with PyTorch 3.8, PyTorch 1.8.1. CUDA 10.2, cuDNN 7.6.5 For installing, follow these instructions:

conda install pytorch=1.8.1 torchvision=0.9.1 -c pytorch
pip install tensorboard einops scikit-image pytorch_msssim opencv-python
conda install pillow

Please use the pillow package downloaded by Conda instead of pip.

Install warmup scheduler:

cd pytorch-gradual-warmup-lr/
python setup.py install
cd ..

For dehazing. Computational complexity: 49.83 GFLOPs total parameters: 5.44M

Train on RESIDE-Indoor

cd ITS
python main.py --mode train --data_dir your_path/reside-indoor

Train on RESIDE-Outdoor

cd OTS
python main.py --mode train --data_dir your_path/reside-outdoor

Evaluation

The pre-trained models are located in the files.

Testing on SOTS-Indoor

cd ITS
python main.py --data_dir your_path/reside-indoor --test_model path_to_its_model

Testing on SOTS-Outdoor

cd OTS
python main.py --data_dir your_path/reside-outdoor --test_model path_to_ots_model

For training and testing, your directory structure should look like this

Your path 
├──reside-indoor
     ├──train
          ├──gt
          └──hazy
     └──test
          ├──gt
          └──hazy
└──reside-outdoor
     ├──train
          ├──gt
          └──hazy
     └──test
          ├──gt
          └──hazy

Citation

If you use our work, please consider citing:

@article{cui2025eenet,
  title={EENet: An effective and efficient network for single image dehazing},
  author={Cui, Yuning and Wang, Qiang and Li, Chaopeng and Ren, Wenqi and Knoll, Alois},
  journal={Pattern Recognition},
  volume={158},
  pages={111074},
  year={2025}
}