.png){kind=icon}
.png){kind=icon}
Return
WETR: Wireframe parsing using deformable Transformers
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cag.2025.104323.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Deformable Transformers improve wireframe parsing with scale-aware attention. • Dynamic attention focuses on line endpoints and junctions for accurate segmentation. • Multi-scale feature refinement enhances robustness to scale variation and occlusion.
Keywords:
Deformable Transformers
Dynamic attention
Line endpoints
Junctions
Multi-scale feature refinement