/ShaDocFormer

[IJCNN 2024] A Transformer-based architecture that is good at detecting and capturing shadow and lighting changes, enabling effective shadow removal.

Primary LanguagePythonMIT LicenseMIT

📋 ShaDocFormer

ShaDocFormer: A Shadow-Attentive Threshold Detector With Cascaded Fusion Refiner for Document Shadow Removal.

Weiwen Chen , Yingtie Lei , Shenghong Luo , Ziyang Zhou , Mingxian Li and Chi-Man Pun

University of Macau, Huizhou University

In International Joint Conference on Neural Networks 2024 (IJCNN 2024)

Paper

⚙️ Usage

Installation

git clone https://github.com/kilito777/ShaDocFormer.git
cd ShaDocFormer
pip install -r requirements.txt

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 Accelerate.

💗 Acknowledgements

This work was supported in part by the Science and Technology Development Fund, Macau SAR, under Grant 0087/2020/A2 and Grant 0141/2023/RIA2.

🛎 Citation

If you find our work helpful for your research, please cite: