.png){kind=icon}
.png){kind=icon}
Return
VTECSeg: Edge-aware hybrid CNN-vision transformer network with zero-shot vision-language-guided region proposals for crack segmentation
.png){kind=icon}
OA
AI
O
X
R
W
T
G
DOI：10.1016/j.autcon.2026.106998.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Zero-shot VLM–SAM workflow for crack patch localization. • Combined text prompts and image encoding for spatial region proposals. • Fusion of vision transformers and edge-aware CNN features. • Pipeline outperforms existing deep learning models. • Supports real-time automated infrastructure inspection.
Keywords:
Crack segmentation
Inspection automation
Deep learning
Computer vision
Vision transformers
Convolutional neural networks
Vision-language models
Region proposals
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