Multimodal Large Language Models (MLLMs) have advanced VQA and now support Vision-DeepResearch systems that use search engines for complex visual-textual fact-finding. However, evaluating these visual and textual search abilities is still difficult, and existing benchmarks have two major limitations. First, existing benchmarks are not visual search-centric: answers that should require visual search are often leaked through cross-textual cues in the text questions or can be inferred from the prior world knowledge in current MLLMs. Second, overly idealized evaluation scenario: On the image-search side, the required information can often be obtained via near-exact matching against the full image, while the text-search side is overly direct and insufficiently challenging. To address these issues, we construct the Vision-DeepResearch benchmark (VDR-Bench) comprising 2,000 VQA instances. All questions are created via a careful, multi-stage curation pipeline and rigorous expert review, designed to assess the behavior of Vision-DeepResearch systems under realistic real-world conditions. Moreover, to address the insufficient visual retrieval capabilities of current MLLMs, we propose a simple multi-round cropped-search workflow. This strategy is shown to effectively improve model performance in realistic visual retrieval scenarios. Overall, our results provide practical guidance for the design of future multimodal deep-research systems. The code will be released in https://github.com/Osilly/Vision-DeepResearch.

翻译：多模态大语言模型（MLLMs）在视觉问答任务上取得了显著进展，并已开始支持基于搜索引擎的视觉深度研究系统，以完成复杂的视觉-文本事实检索任务。然而，评估这些视觉与文本搜索能力仍面临挑战，现有基准存在两大主要局限：其一，现有基准并非以视觉搜索为核心——本需通过视觉搜索获取的答案，往往因文本问题中的跨文本线索而泄露，或可依赖当前MLLMs已有的世界知识推断得出；其二，评估场景过于理想化：在图像搜索方面，所需信息常可通过与完整图像近乎精确的匹配获得，而文本搜索则过于直接且挑战性不足。为应对这些问题，我们构建了视觉深度研究基准（VDR-Bench），包含2000个视觉问答实例。所有问题均通过严谨的多阶段筛选流程与专家评审创建，旨在评估视觉深度研究系统在真实场景下的实际表现。此外，针对当前MLLMs视觉检索能力不足的问题，我们提出一种简单的多轮裁剪搜索工作流程。实验表明，该策略能有效提升模型在真实视觉检索场景中的性能。总体而言，我们的研究结果为未来多模态深度研究系统的设计提供了实用指导。代码将在https://github.com/Osilly/Vision-DeepResearch 发布。