.png){kind=icon}
.png){kind=icon}
Return
Dual-encoder Multi-Scale Refinement network for robust crack segmentation across diverse domains
.png){kind=icon}
OA
AI
H
R
J
R
DOI：10.1016/j.autcon.2026.107004.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Dual-encoder network captures local texture and global crack structure. • CDEM, BRH, and MSF modules preserve thin cracks and sharpen boundaries. • Composite loss improves pixel accuracy and topological consistency. • Strong cross-dataset performance under leave-one-dataset-out testing. • Favorable accuracy-efficiency balance with 11.66M parameters.
Keywords:
Crack segmentation
Deep learning
CNN–ViT hybrid
Multi-scale feature fusion
Boundary refinement
Generalization performance
Infrastructure monitoring
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
11.5
Papers:
6.1K
Citations:
4.2W
Organization
Cited Papers