Row-Column Separated Attention Based Low-Light Image/Video Enhancement
Chengqi Dong 1 , Zhiyuan Cao 2 , Tuoshi Qi 1 , Kexin Wu 1 , Yixing Gao 1 , and Fan Tang 3 1 School of Artificial Intelligence, Jilin University, China 2 College of Software, Jilin University, China 3 Institute of Computing Technology, Chinese Academy of Sciences, China · 2026
Read Row-Column Separated Attention Based Low-Light Image/Video Enhancement by Chengqi Dong 1 , Zhiyuan Cao 2 , Tuoshi Qi 1 , Kexin Wu 1 , Yixing Gao 1 , and Fan Tang 3 1 School of Artificial Intelligence, Jilin University, China 2 College of Software, Jilin University, China 3 Institute of Computing Technology, Chinese Academy of Sciences, China (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-07428.txt.
Description
Low-light photos and videos suffer noise and lost detail, and usual attention mechanisms that could supply global context are too heavy for the job. Chengqi Dong and five co-authors insert a Row-Column Separated Attention module after an improved U-Net, summarizing each feature map by row and column means and maxima.
That compact summary lets global information guide local enhancement with far fewer settings, and two new time loss functions extend the method to video while keeping frames consistent. Experiments on standard low-light image and video data sets confirm the gains. Computer vision builders get a lightweight enhancement module that brightens dark footage without blowing up the model.
Catalog record
| CREATOR | Chengqi Dong 1 , Zhiyuan Cao 2 , Tuoshi Qi 1 , Kexin Wu 1 , Yixing Gao 1 , and Fan Tang 3 1 School of Artificial Intelligence, Jilin University, China 2 College of Software, Jilin University, China 3 Institute of Computing Technology, Chinese Academy of Sciences, China |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-07428