1
Return

FreedomDiVe: Task-free image fusion via marginal distribution-based diffusion variational estimation

delete2026-07-07
delete0
PRE
AI
Z
Zihan Gui
H
He Huang
G
Guangyi Yang
J
Jun Liu *
J
Jun Ma
W
Wei He *
DOI：10.1016/j.patcog.2026.114400delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Diffuse images into Gaussian space for fusion to enhance general applicability. • Construct a universal prior to address the lack of ground truth in image fusion. • We propose FreedomDiVE, a task-free image fusion method via Diffusion Models. • Experiments across diverse fusion tasks demonstrate FreedomDiVE’s superiority.
Keywords:
Task-agnostic image fusion
Diffusion model
Score based model
Variational estimation

Journal

Pattern Recognition cover
Pattern Recognition
IF:
7.6
Papers:
1.3W
Citations:
4.5W

Organization

T
The University of Hong Kong
Scholars:
5.7K
Papers: 2.8K
Citations: 7
W
wuhan university
Scholars:
7.8W
Papers: 5.7W
Citations: 70
Cited Papers

Cited Papers

Citing Papers

Citing Papers