AdaGround：Token 归一化自适应缩放加验证器在环精调的无训练 GUI grounding 框架

View PDF HTML (experimental)

Abstract:GUI grounding on high-resolution screens fails mostly on small targets, and zooming in is a common training-free remedy. However, most existing zooming pipelines do not match the zoom to the target's size, accept the zoomed answer unconditionally and never revisit a wrong answer. To address these issues, we propose AdaGround, a training-free framework with three components: (1) Token-Normalized Adaptive Zooming, which sets the magnification, field of view and zoom depth of each sample from the model's own estimate of the target size; (2) Verifier-in-the-Loop Refinement, which uses one marked-click Yes/No verifier to keep or reject each zoom, choose between zoom chains, trigger re-grounding and, with an existence query, decide when to abstain; and (3) Compositional Structure Grounding, which composes spreadsheet cells from their headers. On Qwen3-VL models, AdaGround surpasses state-of-the-art training-free zooming methods and many trained GUI models, improving accuracy over the backbone by 7.4 to 20.6 points on OSWorld-G and ScreenSpot-Pro, and its gains carry over to held-out UI-Vision. Code is available at this https URL

Comments:	24 pages, 5 figures, 14 tables
Subjects:	Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
ACM classes:	I.2.10; H.5.2
Cite as:	arXiv:2610.36874 [cs.CV]
	(or arXiv:2610.36874v1 [cs.CV] for this version)
	https://doi.org/10.48550/arXiv.2610.36874 arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Bingchao Wang [view email]
[v1] Tue, 27 Oct 2026 09:12:35 UTC (478 KB)