1
Return

KANFuse: Enhancing infrared and visible image fusion through nonlinear representation modeling

delete2025-12-15
delete0
PRE
AI
Y
Yongzi Zhang
S
S.G. Li *
A
A. C. Fang
X
Xinglong He
D
Daoheng Zhu
K
Keer Wu
X
Xiuchun Xiao *
DOI：10.1016/j.infrared.2025.106319delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Tok-KAN boosts nonlinear feature extraction and enhances representation flexibility. • WCBs retain edges and textures while adaptively suppressing noise. • DFM uses channel and large kernel attention for dynamic multi-modal fusion. • SFL guides color retention from visible images in second-phase training.

Journal

I
Infrared Physics and Technology
IF:
3.4
Papers:
5.8K
Citations:
1.2W

Organization

G
Guangdong Ocean University
Scholars:
6.6K
Papers: 3.7K
Citations: 4.8K
Cited Papers

Cited Papers

Citing Papers

Citing Papers