arXiv 2605.130342026-05-13cs.CVcs.IR

ViDR：基于源视觉证据的多模态深度研究报告 grounding

ViDR: Grounding Multimodal Deep Research Reports in Source Visual Evidence

School of Software and Microelectronics, Peking University（北京大学软件与微电子学院）；National Key Laboratory of Data Space Technology and System（数据空间技术与系统国家重点实验室）；School of Software Engineering, Beijing Jiaotong University（北京交通大学软件学院）；College of Computer Science and Electronic Engineering, Hunan University（湖南大学计算机科学与电子工程学院）

Zhuofan Shi, Peilun Jia, Baoqin Sun, Haiyang Shen, Sixiong Xie, Yun Ma, Xiang Jing

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

ViDR通过源视觉证据构建多模态深度研究报告框架，提升报告质量与证据可验证性。

摘要：

最近的深度研究系统通过迭代检索和推理提升了大语言模型生成长报告的能力。然而，大多数以文本为中心的系统主要依赖文本证据，而多模态系统通常只能弱化检索图像或自行生成图表，导致源图未被充分利用。本文提出ViDR，一种多模态深度研究框架，将长报告基于源图证据。ViDR将源图视为可检索、可解释、可路由和可验证的证据对象，同时在需要时生成分析图表。它构建了连接主张到文本和视觉证据的证据索引大纲，通过上下文感知过滤、大纲感知重排序和基于VLM的视觉分析，将嘈杂的网络图像转化为源图证据原子，并为每个部分生成特定的证据。ViDR进一步验证视觉参考以减少幻觉或错误放置的图像。我们还引入了MMR Bench+，一个评估深度研究报告中视觉证据使用的基准，涵盖源图检索、放置、解释、可验证性和分析图表生成。实验表明，ViDR在整体报告质量、源图整合和可验证性方面优于强大的商业和开源基线。这些结果表明，源视觉证据对多模态深度研究至关重要，因为它增强了证据基础、视觉支持和报告可验证性。

英文摘要：

Recent deep research systems have improved the ability of large language models to produce long, grounded reports through iterative retrieval and reasoning. However, most text-centered systems rely mainly on textual evidence, while multimodal systems often retrieve images only weakly or generate charts themselves, leaving source figures underused as evidence. We present ViDR, a multimodal deep research framework that grounds long-form reports in source figures. ViDR treats source figures as retrievable, interpretable, routable, and verifiable evidence objects, while still generating analytical charts when needed. It builds an evidence-indexed outline linking claims to textual and visual evidence, refines noisy web images into source-figure evidence atoms through context-aware filtering, outline-aware reranking, and VLM-based visual analysis, and generates each section with section-specific evidence. ViDR further validates visual references to reduce hallucinated or misplaced figures. We also introduce MMR Bench+, a benchmark for evaluating visual evidence use in deep research reports, covering source-figure retrieval, placement, interpretation, verifiability, and analytical chart generation. Experiments show that ViDR improves overall report quality, source-figure integration, and verifiability over strong commercial and open-source baselines. These results suggest that source visual evidence is important for multimodal deep research, as it strengthens evidential grounding, visual support, and report verifiability.

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。