UI-Zoomer: Uncertainty-Driven Adaptive Zoom-In for GUI Grounding
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913035740774400 |
|---|---|
| author | Tang, Fei Chen, Bofan Lu, Zhengxi Chen, Tongbo Nong, Songqin Jiang, Tao Xu, Wenhao Lu, Weiming Xiao, Jun Zhuang, Yueting Shen, Yongliang |
| author_facet | Tang, Fei Chen, Bofan Lu, Zhengxi Chen, Tongbo Nong, Songqin Jiang, Tao Xu, Wenhao Lu, Weiming Xiao, Jun Zhuang, Yueting Shen, Yongliang |
| contents | GUI grounding, which localizes interface elements from screenshots given natural language queries, remains challenging for small icons and dense layouts. Test-time zoom-in methods improve localization by cropping and re-running inference at higher resolution, but apply cropping uniformly across all instances with fixed crop sizes, ignoring whether the model is actually uncertain on each case. We propose \textbf{UI-Zoomer}, a training-free adaptive zoom-in framework that treats both the trigger and scale of zoom-in as a prediction uncertainty quantification problem. A confidence-aware gate fuses spatial consensus among stochastic candidates with token-level generation confidence to selectively trigger zoom-in only when localization is uncertain. When triggered, an uncertainty-driven crop sizing module decomposes prediction variance into inter-sample positional spread and intra-sample box extent, deriving a per-instance crop radius via the law of total variance. Extensive experiments on ScreenSpot-Pro, UI-Vision, and ScreenSpot-v2 demonstrate consistent improvements over strong baselines across multiple model architectures, achieving gains of up to +13.4\%, +10.3\%, and +4.2\% respectively, with no additional training required. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_14113 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | UI-Zoomer: Uncertainty-Driven Adaptive Zoom-In for GUI Grounding Tang, Fei Chen, Bofan Lu, Zhengxi Chen, Tongbo Nong, Songqin Jiang, Tao Xu, Wenhao Lu, Weiming Xiao, Jun Zhuang, Yueting Shen, Yongliang Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language GUI grounding, which localizes interface elements from screenshots given natural language queries, remains challenging for small icons and dense layouts. Test-time zoom-in methods improve localization by cropping and re-running inference at higher resolution, but apply cropping uniformly across all instances with fixed crop sizes, ignoring whether the model is actually uncertain on each case. We propose \textbf{UI-Zoomer}, a training-free adaptive zoom-in framework that treats both the trigger and scale of zoom-in as a prediction uncertainty quantification problem. A confidence-aware gate fuses spatial consensus among stochastic candidates with token-level generation confidence to selectively trigger zoom-in only when localization is uncertain. When triggered, an uncertainty-driven crop sizing module decomposes prediction variance into inter-sample positional spread and intra-sample box extent, deriving a per-instance crop radius via the law of total variance. Extensive experiments on ScreenSpot-Pro, UI-Vision, and ScreenSpot-v2 demonstrate consistent improvements over strong baselines across multiple model architectures, achieving gains of up to +13.4\%, +10.3\%, and +4.2\% respectively, with no additional training required. |
| title | UI-Zoomer: Uncertainty-Driven Adaptive Zoom-In for GUI Grounding |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2604.14113 |