UltraVR：面向证据推理的诊断性超分辨率图像VQA基准
UltraVR: A Diagnostic Ultra-Resolution Image-VQA Benchmark for Evidence-Grounded Reasoning
AI总结：
提出UltraVR基准，通过结构化思维链标注诊断视觉语言模型在超分辨率图像上的证据推理能力，发现模型在证据定位和局部感知环节错误集中。
摘要：
视觉语言模型（VLM）在视觉问答和多模态推理基准上表现出色。然而，它们在超分辨率图像上的能力——其中关键证据微小、细微、空间遥远或分布广泛——仍不清楚。现有评估主要报告最终答案准确率，对模型是否获取并整合必要视觉证据的洞察有限。我们引入UltraVR，一个面向超分辨率图像上基于证据的视觉推理的诊断性基准。UltraVR涵盖四个高价值场景：CCTV监控、遥感（RS）、全切片图像（WSI）病理学和工业异常检测（AD）。这些领域提出互补挑战：拥挤CCTV场景中的细粒度目标定位、RS中的长程空间比较、WSI中的多尺度证据导航以及重复工业布局中的细微不规则检测。除了标准QA三元组，每个实例包括一个结构化的真实思维链，包含步骤级问题、中间答案和推理标签。这些标签将推理分解为证据定位、局部感知、量化、证据整合和决策推断，从而实现对黑盒评分的流程级诊断。使用UltraVR，我们评估前沿VLM，并表明当前模型在超分辨率推理上仍远不可靠。重要的是，结构化注释使我们能够定位从视觉到决策流水线中的失败：错误集中在证据定位和局部感知，而当提供中间视觉事实时，下游推理通常能够恢复。这些发现表明UltraVR是一个诊断性测试平台，不仅衡量VLM是否回答正确，还衡量其超分辨率推理过程在何处中断。
英文摘要：
Vision-language models (VLMs) excel on visual question answering and multimodal reasoning benchmarks. Yet their capability on ultra-resolution images - where critical evidence is tiny, subtle, spatially distant, or distributed - remains unclear. Existing evaluations largely report final-answer accuracy, offering limited insight into whether models acquire and integrate the necessary visual evidence. We introduce UltraVR, a diagnostic benchmark for evidence-grounded visual reasoning over ultra-resolution images. UltraVR spans four high-value scenarios: CCTV surveillance, remote sensing (RS), whole-slide image (WSI) pathology, and industrial anomaly detection (AD). These domains pose complementary challenges: fine-grained object grounding in crowded CCTV scenes, long-range spatial comparison in RS, multi-scale evidence navigation in WSI, and subtle irregularity detection in repetitive industrial layouts. Beyond standard QA triples, each instance includes a structured ground-truth chain of thought with step-level questions, intermediate answers, and reasoning labels. These labels decompose reasoning into evidence grounding, local perception, quantification, evidence integration, and decision inference, enabling process-level diagnosis over black-box scoring. Using UltraVR, we evaluate frontier VLMs and show that current models remain far from reliable on ultra-resolution reasoning. Importantly, the structured annotations allow us to localize failures across the visual-to-decision pipeline: errors concentrate in evidence grounding and local perception, while downstream inference often recovers when intermediate visual facts are supplied. These findings demonstrate UltraVR as a diagnostic testbed for measuring not only whether VLMs answer correctly, but where their ultra-resolution reasoning process breaks.
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。
