🌧🌫❄WeatherBench: A Real-World Benchmark Dataset for All-in-One Adverse Weather Image Restoration [ACM MM 2025 Datasets Track]
Qiyuan Guan* 1, Qianfeng Yang* 1, Xiang Chen* 2, Tianyu Song 3, Guiyue Jin 1, Jiyu Jin 1
Dalian Polytechnic University1, Nanjing University of Science and Technology2, Dalian Martime University3
[Paper]
👉️ 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/
- ✔ Release the dataset
- ✔ Release the visual results
| Download Link | Description |
|---|---|
| Google Drive / Baidu Netdisk (dlpu) | Tran: 41,402 pairs. Test: 600 pairs. |
We provide evaluation code (Python Code) to assess PSNR, SSIM, and LPIPS results. You need to set the parameters --generated_images_path, --target_path, and --Score_save_path.
If WeatherBench is helpful for you, please help star the GitHub Repo. Thanks!
| Method | Download Link |
|---|---|
| DehazeFormer | Google Drive / Baidu Netdisk (deha) |
| DCMPNet | Google Drive / Baidu Netdisk (dcmp) |
| DRSformer | Google Drive / Baidu Netdisk (drsf) |
| NeRD-Rain | Google Drive / Baidu Netdisk (nerd) |
| SnowFormer | Google Drive / Baidu Netdisk (snof) |
| MPRNet | Google Drive / Baidu Netdisk (mprn) |
| Restormer | Google Drive / Baidu Netdisk (rest) |
| AirNet | Google Drive / Baidu Netdisk (airn) |
| TransWeather | Google Drive / Baidu Netdisk (tran) |
| PromptIR | Google Drive / Baidu Netdisk (prom) |
| WGWS-Net | Google Drive / Baidu Netdisk (wgws) |
| DiffUIR | Google Drive / Baidu Netdisk (diff) |
| MWFormer | Google Drive / Baidu Netdisk (mwfo) |
| Histoformer | Google Drive / Baidu Netdisk (hist) |
| AdaIR | Google Drive / Baidu Netdisk (adai) |
If this work is helpful for your research, please consider citing the following BibTeX entry.
@inproceedings{guan2025weatherbench,
title={Weatherbench: A real-world benchmark dataset for all-in-one adverse weather image restoration},
author={Guan, Qiyuan and Yang, Qianfeng and Chen, Xiang and Song, Tianyu and Jin, Guiyue and Jin, Jiyu},
booktitle={Proceedings of the 33rd ACM International Conference on Multimedia},
pages={12607--12613},
year={2025}
}
If you have any questions, please feel free to contact qyuanguan@gmail.com or csqianfengyang@163.com.