Improving GUI Grounding with Explicit Position-to-Coordinate Mapping

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Suyuchen, Zhang, Tianyu, Masry, Ahmed, Pal, Christopher, Gella, Spandana, Liu, Bang, Taslakian, Perouz
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911190957948928
author Wang, Suyuchen
Zhang, Tianyu
Masry, Ahmed
Pal, Christopher
Gella, Spandana
Liu, Bang
Taslakian, Perouz
author_facet Wang, Suyuchen
Zhang, Tianyu
Masry, Ahmed
Pal, Christopher
Gella, Spandana
Liu, Bang
Taslakian, Perouz
contents GUI grounding, the task of mapping natural-language instructions to pixel coordinates, is crucial for autonomous agents, yet remains difficult for current VLMs. The core bottleneck is reliable patch-to-pixel mapping, which breaks when extrapolating to high-resolution displays unseen during training. Current approaches generate coordinates as text tokens directly from visual features, forcing the model to infer complex position-to-pixel mappings implicitly; as a result, accuracy degrades and failures proliferate on new resolutions. We address this with two complementary innovations. First, RULER tokens serve as explicit coordinate markers, letting the model reference positions similar to gridlines on a map and adjust rather than generate coordinates from scratch. Second, Interleaved MRoPE (I-MRoPE) improves spatial encoding by ensuring that width and height dimensions are represented equally, addressing the asymmetry of standard positional schemes. Experiments on ScreenSpot, ScreenSpot-V2, and ScreenSpot-Pro show consistent gains in grounding accuracy, with the largest improvements on high-resolution interfaces. By providing explicit spatial guidance rather than relying on implicit learning, our approach enables more reliable GUI automation across diverse resolutions and platforms.
format Preprint
id arxiv_https___arxiv_org_abs_2510_03230
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Improving GUI Grounding with Explicit Position-to-Coordinate Mapping
Wang, Suyuchen
Zhang, Tianyu
Masry, Ahmed
Pal, Christopher
Gella, Spandana
Liu, Bang
Taslakian, Perouz
Computer Vision and Pattern Recognition
Artificial Intelligence
GUI grounding, the task of mapping natural-language instructions to pixel coordinates, is crucial for autonomous agents, yet remains difficult for current VLMs. The core bottleneck is reliable patch-to-pixel mapping, which breaks when extrapolating to high-resolution displays unseen during training. Current approaches generate coordinates as text tokens directly from visual features, forcing the model to infer complex position-to-pixel mappings implicitly; as a result, accuracy degrades and failures proliferate on new resolutions. We address this with two complementary innovations. First, RULER tokens serve as explicit coordinate markers, letting the model reference positions similar to gridlines on a map and adjust rather than generate coordinates from scratch. Second, Interleaved MRoPE (I-MRoPE) improves spatial encoding by ensuring that width and height dimensions are represented equally, addressing the asymmetry of standard positional schemes. Experiments on ScreenSpot, ScreenSpot-V2, and ScreenSpot-Pro show consistent gains in grounding accuracy, with the largest improvements on high-resolution interfaces. By providing explicit spatial guidance rather than relying on implicit learning, our approach enables more reliable GUI automation across diverse resolutions and platforms.
title Improving GUI Grounding with Explicit Position-to-Coordinate Mapping
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2510.03230