EVSSM
[CVPR 2025] Efficient Visual State Space Model for Image Deblurring; 1st place on AIM 2025 challenge on High FPS Motion Deblurring:
Install / Use
/learn @kkkls/EVSSMREADME
Efficient Visual State Space Model for Image Deblurring
Paper: https://openaccess.thecvf.com/content/CVPR2025/papers/Kong_Efficient_Visual_State_Space_Model_for_Image_Deblurring_CVPR_2025_paper.pdf
Dependencies
conda env create -f environment.yml
Train
bash train.sh
Test
bash test.sh
checkpoint
GoPro and Realblur:(https://drive.google.com/drive/folders/1_G0hf1KX_SAwyK41bdGm2q0q63B6wN40?usp=drive_link)
BibTeX
@inproceedings{EVSSM,
title={Efficient Visual State Space Model for Image Deblurring},
author={Kong, Lingshun and Dong, Jiangxin and Tang, Jinhui and Yang, Ming-Hsuan and Pan, Jinshan},
booktitle={CVPR},
year={2025}
}
