Rethinking Multi-Scale Representations in Deep Deraining Transformer (AAAI2024)

Welcome to visit our website (专注底层视觉领域的信息服务平台) for low-level vision: https://lowlevelcv.com/

🛠️ Training and Testing

1. Please put datasets in the folder Datasets/.
2. Follow the instructions below to begin training our model.

bash train.sh

Run the script then you can find the generated experimental logs in the folder checkpoints.

3. Follow the instructions below to begin testing our model.

python test.py

Run the script then you can find the output visual results in the folder results/.

🤖 Pre-trained Models

Models	MSDT
Rain200L	Google Drive / Baidu Netdisk (8ajd)
Rain200H	Google Drive / Baidu Netdisk (97lm)
DID-Data	Google Drive / Baidu Netdisk (5g4p)
DDN-Data	Google Drive / Baidu Netdisk (b0b5)
SPA-Data	Google Drive / Baidu Netdisk (x0i5)

🚨 Performance Evaluation

See folder "evaluations"

1. for Rain200L/H and SPA-Data datasets: PSNR and SSIM results are computed by using this Matlab Code.

2. for DID-Data and DDN-Data datasets: PSNR and SSIM results are computed by using this Matlab Code.

🚀 Visual Deraining Results

Methods	MSDT
Rain200L	Baidu Netdisk (1xkc)
Rain200H	Baidu Netdisk (yr3n)
DID-Data	Baidu Netdisk (242e)
DDN-Data	Baidu Netdisk (2pwk)
SPA-Data	Baidu Netdisk (cag0)

👍 Acknowledgement

Thanks for their awesome works (DeepRFT and DRSformer).

📘 Citation

Please consider citing our work as follows if it is helpful.

@inproceedings{chen2024rethinking,
  title={Rethinking Multi-Scale Representations in Deep Deraining Transformer},
  author={Chen, Hongming and Chen, Xiang and Lu, Jiyang and Li, Yufeng},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={38},
  number={2},
  pages={1046--1053},
  year={2024}
}