Skip to content

[WACV 2025] High-Fidelity Document Stain Removal via A Large-Scale Real-World Dataset and A Memory-Augmented Transformer

License

Notifications You must be signed in to change notification settings

CXH-Research/StainRestorer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Oct 31, 2024
d412f86 ยท Oct 31, 2024

History

5 Commits
Oct 29, 2024
Oct 29, 2024
Oct 29, 2024
Oct 29, 2024
Oct 29, 2024
Oct 29, 2024
Oct 31, 2024
Oct 29, 2024
Oct 29, 2024
Oct 29, 2024

Repository files navigation

StainRestorer

High-Fidelity Document Stain Removal via A Large-Scale Real-World Dataset and A Memory-Augmented Transformer

Mingxian Li ๐Ÿ‘จโ€๐Ÿ’ปโ€ , Hao Sun ๐Ÿ‘จโ€๐Ÿ’ปโ€ , Yingtie Lei ๐Ÿ‘จโ€๐Ÿ’ปโ€ , Xiaofeng Zhang , Yihang Dong , Yilin Zhou , Zimeng Li , Xuhang Chen ๐Ÿ“ฎ ( ๐Ÿ‘จโ€๐Ÿ’ปโ€ Equal contributions, ๐Ÿ“ฎ Corresponding author)

Huizhou Univeristy, University of Macau, Shanghai Jiao Tong University, SIAT CAS, Shenzhen Polytechnic University

In IEEE/CVF Winter Conference on Applications of Computer Vision 2025 (WACV 2025)

๐Ÿ”ฎ Dataset

Kaggle

StainDoc is the first large-scale high-resolution dataset that includes ground truth data specifically for the task of document stain removal.

StainDoc_mark and StainDoc_seal are made with the process in DocDiff.

โš™๏ธ 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 Accelerate.

Inference

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

python infer.py

Citation

About

[WACV 2025] High-Fidelity Document Stain Removal via A Large-Scale Real-World Dataset and A Memory-Augmented Transformer

Topics

Resources

License

Stars

Watchers

Forks

Languages