LoFormer: Local Frequency Transformer for Image Deblurring (ACM MM 24)

Xintian Mao, Jiansheng Wang, Xingran Xie , Qingli Li and Wang Yan

Quick Run

Training

1. Download GoPro training and testing data
2. To train the LoFormer, run

   cd LoFormer
   ./train_8gpu.sh Motion_Deblurring/Options/train_GoPro_LoformerLarge_600k_8gpu.yml

Evaluation

To test the pre-trained models of Deblur 百度网盘(提取码:ca3l) or GoogleDrive on your own images, run

python Motion_Deblurring/val.py 

Core Code

FreqLC can be seen in LoFormer/FreqLC

Results

Results on GoPro, HIDE, Realblur test sets: 百度网盘(提取码:rcx5)

Citation

If you use , please consider citing:

@inproceedings{xintm2024LoFormer, 
    title = {LoFormer: Local Frequency Transformer for Image Deblurring},
    author = {Xintian Mao, Jiansheng Wang, Xingran Xie, Qingli Li and Yan Wang}, 
    booktitle = {Proc. ACM MM}, 
    year = {2024}
    }

Contact

If you have any question, please contact mxt_invoker1997@163.com

Our Related Works

• Deep Residual Fourier Transformation for Single Image Deblurring, arXiv 2021. Paper | Code
• Intriguing Findings of Frequency Selection for Image Deblurring, AAAI 2023. Paper | Code
• AdaRevD: Adaptive Patch Exiting Reversible Decoder Pushes the Limit of Image Deblurring, CVPR 2024. Paper | Code

Reference Code:

• https://github.com/megvii-research/NAFNet
• https://github.com/INVOKERer/DeepRFT/tree/AAAI2023
• https://github.com/swz30/Restormer

Acknowledgment: This code is based on the BasicSR toolbox.