Paper Overview
- Field: Computer Vision (GUI grounding / GUI agents)
- Authors: Borui Zhang, Bo Zhang, Bo Wang, Wenzhao Zheng et al.
- Published: 2026-05-07
- arXiv: 2605.06664
- Code: https://github.com/Neur-IO/BAMI
- Coarse-to-fine focusing
- Candidate selection
- Paper: https://arxiv.org/abs/2605.06664
- Code: https://github.com/Neur-IO/BAMI
Summary (translated from the forum post)
GUI grounding is a key capability that enables GUI agents to perform tasks such as clicking and dragging. However, existing models often underperform on complex-scene benchmarks like ScreenSpot-Pro.
Using the authors' proposed Mask Prediction Distribution (MPD) attribution method, they identify two main sources of error:
1. High image resolution — leading to *precision bias*. 2. Complex interface elements — leading to *ambiguity bias*.
To address these issues, the paper proposes Bias-Aware Manipulation Inference (BAMI), which includes two key operations:
Extensive experimental results show that BAMI significantly improves the accuracy of various GUI grounding models in a training-free setting. For example, applying BAMI to the TianXi-Action-7B model raises its ScreenSpot-Pro accuracy from 51.9% to 57.8%. Ablation studies further confirm BAMI's robustness across different parameter configurations, highlighting its stability and effectiveness.