Referring flexible image restoration
Lists
Only you see this
Abstract
Highlights•A novel task: Referring Flexible Image Restoration, prompt-based degrade-removal.•A novel dataset called RFIR, containing 153,423 samples with five degradation types.•A multi-task model for specific degradation-removal and multi-deg classification.•A highly efficient attention module called MHASA with linear complexity.•A novel fusion method called MHACA with better performance than vanilla MHCA.
BibTeX
@article{guan2025referring,
author = {Runwei Guan and Rongsheng Hu and Zhuhao Zhou and Tianlang Xue and Ka Lok Man and Jeremy S. Smith and Eng Gee Lim and Weiping Ding and Yutao Yue},
title = {{Referring flexible image restoration}},
journal = {Expert Systems with Applications},
year = {2025},
doi = {10.1016/j.eswa.2025.126857},
}