MVP：多视图预测提升 GUI 视觉定位
MVP: Multiple View Prediction Improves GUI Grounding
对做视觉定位/VLM 的研究者，MVP 提供了一种免训练的多视图推理与聚类集成策略：用指令到图像的注意力分数生成多样化裁剪视图，再对多视图坐标预测做最密空间簇质心聚合，可直接套用到任何坐标回归式 VLM 上以抑制单视图预测的不稳定性。
原文依据 we propose Multi-View Prediction (MVP), a training-free framework that enhances grounding performance through multi-view inference
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论GUI 视觉定位将自然语言指令转化为精确的像素坐标，是构建实用 GUI 智能体的关键。然而我们观察到，现有视觉定位模型存在显著的坐标预测不稳定性——微小的视觉扰动（例如裁剪掉几个像素）就可能大幅改变预测结果，使原本正确的结果翻转为错误。这种不稳定性严重损害了模型性能，对高分辨率和小尺寸 UI 元素的样本尤为明显。为解决该问题，我们提出多视图预测（MVP），一个免训练框架，通过多视图推理提升视觉定位性能。我们的关键洞见是：单视图预测可能不稳定，但聚合来自多个精心裁剪视图的预测，能够有效区分正确坐标与离群值。MVP 包含两个组件：(1) 注意力引导的视图提议，依据指令到图像的注意力分数导出多样化视图；(2) 多坐标聚类，通过选取最密集空间簇的质心来集成预测。大量实验表明 MVP 在多种模型与基准上均有效。值得注意的是，在 ScreenSpot-Pro 上，MVP 将 UI-TARS-1.5-7B 提升至 56.1%，GTA1-7B 提升至 61.7%，Qwen3VL-8B-Instruct 提升至 65.3%，Qwen3VL-32B-Instruct 提升至 74.0%。
英文原文摘要 1369 字符
GUI grounding, which translates natural language instructions into precise pixel coordinates, is essential for developing practical GUI agents. However, we observe that existing grounding models exhibit significant coordinate prediction instability--minor visual perturbations (e.g., cropping a few pixels) can drastically alter predictions, flipping results between correct and incorrect. This instability severely undermines model performance, especially for samples with high-resolution and small UI elements. To address this issue, we propose Multi-View Prediction (MVP), a training-free framework that enhances grounding performance through multi-view inference. Our key insight is that while single-view predictions may be unstable, aggregating predictions from multiple carefully cropped views can effectively distinguish correct coordinates from outliers. MVP comprises two components: (1) Attention-Guided View Proposal, which derives diverse views guided by instruction-to-image attention scores, and (2) Multi-Coordinates Clustering, which ensembles predictions by selecting the centroid of the densest spatial cluster. Extensive experiments demonstrate MVP's effectiveness across various models and benchmarks. Notably, on ScreenSpot-Pro, MVP boosts UI-TARS-1.5-7B to 56.1%, GTA1-7B to 61.7%, Qwen3VL-8B-Instruct to 65.3%, and Qwen3VL-32B-Instruct to 74.0%.