\ extsc{GUI-Spotlight}：面向增强型GUI视觉定位的自适应迭代聚焦细化方法
\textsc{GUI-Spotlight}: Adaptive Iterative Focus Refinement for Enhanced GUI Visual Grounding
AI总结：
针对GUI系统视觉定位可靠性不足、难以执行精准指针操作的问题，本研究提出GUI-Spotlight模型，通过动态调用专用工具迭代缩小屏幕关注区域，仅用少量训练样本就在ScreenSpot-Pro基准上超越了使用更大训练集的同类模型。
摘要：
多模态大语言模型（MLLMs）显著拓展了图形用户界面（GUI）系统的能力，推动其从受控模拟环境进入跨多平台的复杂真实世界场景。然而，实际效用仍受限于视觉定位的可靠性，也就是将文本参考映射到屏幕上确切元素的能力，这一局限导致系统无法准确执行点击、拖拽等指针级操作。为解决该问题，我们提出GUI-Spotlight——一种专为图像支撑推理训练的模型，它能动态调用多个专用工具，迭代式地将注意力聚焦到屏幕的相关区域，从而大幅提升视觉定位准确率。在ScreenSpot-Pro基准测试上，仅用18.5K训练样本训练的GUI-Spotlight达到了52.8%的准确率，超过了V2P-7B（使用9.6M训练样本时准确率为50.6%）和GTA-1-7B（使用1.56M训练样本时准确率为50.1%）。
英文摘要：
Multimodal large language models (MLLMs) have markedly expanded the competence of graphical user-interface (GUI) systems, propelling them beyond controlled simulations into complex, real-world environments across diverse platforms. However, practical usefulness is still bounded by the reliability of visual grounding, i.e., mapping textual references to exact on-screen elements. This limitation prevents the system from accurately performing pointer-level actions such as clicking or dragging. To address it, we introduce GUI-Spotlight -- a model trained for image-grounded reasoning that dynamically invokes multiple specialized tools to iteratively narrow its focus to the relevant region of the screen, thereby substantially improving visual grounding accuracy. On the ScreenSpot-Pro benchmark, GUI-Spotlight trained with only 18.5K training samples achieves 52.8\% accuracy, surpassing V2P-7B (50.6\% with 9.6M training samples) and GTA-1-7B (50.1\% with 1.56M training samples).
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
