MRD：面向高分辨率图像理解的多分辨率检索-检测融合
MRD: Multi-resolution Retrieval-Detection Fusion for High-Resolution Image Understanding
把开放词表检测器当作定位先验，与多分辨率裁剪检索做免训练融合，为 MLLM 处理高分辨率图像提供了一条可直接套用的「检索+检测」双路增强管线，可迁移到任意现成 MLLM 上。
原文依据 it integrates open-vocabulary object detection (OVD) as localization priors within a unified framework
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论理解高分辨率（HR）图像仍是多模态大语言模型（MLLM）面临的关键挑战。近期方法借助基于视觉的检索增强生成（RAG）从高分辨率图像中检索与查询相关的裁剪区域，从而提升 MLLM 的理解能力。然而，这一范式往往导致目标碎片化，产生语义偏差与不完整检索，同时还会从无关背景块中引入误检。为解决这些问题，我们提出多分辨率检索-检测（MRD），一个免训练框架，从局部与全局两个视角增强高分辨率图像理解。在局部层面，MRD 通过多分辨率语义融合强制跨尺度语义一致性，以缓解单一分辨率偏差并减轻目标碎片化。在全局层面，它将开放词表目标检测（OVD）作为定位先验整合进统一框架。在多个 MLLM 上于高分辨率图像基准开展的大量实验表明，MRD 在单目标与多目标理解任务上均达到最优性能。代码将发布于：https://github.com/yf0412/MRD。
英文原文摘要 1185 字符
Understanding high-resolution (HR) images remains a critical challenge for multimodal large language models (MLLMs). Recent approaches leverage vision-based retrieval-augmented generation (RAG) to retrieve query-relevant crops from HR images, improving understanding capacity of MLLMs. However, this paradigm often leads to object fragmentation, resulting in semantic bias and incomplete retrieval, while also introducing false positives from irrelevant background patches. To address these issues, we propose Multi-resolution Retrieval-Detection (MRD), a training-free framework that enhances HR image understanding from both local and global perspectives. Locally, MRD enforces cross-scale semantic consistency via multi-resolution semantic fusion to mitigate single-resolution bias and alleviate object fragmentation. Globally, it integrates open-vocabulary object detection (OVD) as localization priors within a unified framework. Extensive experiments across multiple MLLMs on HR image benchmarks demonstrate that MRD achieves state-of-the-art (SOTA) performance on both single-object and multi-object understanding tasks. Code will be available at: https://github.com/yf0412/MRD.