Multimodal large language models (MLLMs) often miss small details and spatial relations in cluttered scenes, leading to errors in fine-grained perceptual grounding. We introduce AttWarp, a lightweight method that allocates more resolution to query-relevant content while compressing less informative areas, all while preserving global context. At test time, the approach uses an MLLM's cross-modal attention to perform rectilinear warping of the input image, reallocating spatial resolution toward regions the model deems important, without changing model weights or architecture. This attention-guided warping preserves all original image information but redistributes it non-uniformly, so small objects and subtle relationships become easier for the same model to read while the global layout remains intact. Across five benchmarks (TextVQA, GQA, DocVQA, POPE, MMMU) and four MLLMs (LLaVA, Qwen-VL, InternVL, and InstructBLIP), AttWarp consistently improves accuracy, strengthens compositional reasoning, and reduces hallucinations, outperforming four competitive baselines that manipulate raw images at test time. Together, these results show that attention-guided warping prioritizes information relevant to the query while preserving context, and that the same MLLMs perform better when given such warped inputs.

翻译：多模态大语言模型（MLLMs）在处理杂乱场景时，常常遗漏微小细节与空间关系，导致细粒度感知定位出现错误。我们提出AttWarp，一种轻量级方法，它在保持全局上下文的同时，将更多分辨率分配给查询相关的内容，并压缩信息量较少的区域。该方法在测试时利用MLLM的跨模态注意力对输入图像进行直线扭曲，在不改变模型权重或架构的情况下，将空间分辨率重新分配给模型认为重要的区域。这种注意力引导的扭曲保留了所有原始图像信息，但对其进行非均匀重新分布，从而使同一模型更容易识别小物体和细微关系，同时保持全局布局不变。在五个基准测试（TextVQA、GQA、DocVQA、POPE、MMMU）和四种MLLM（LLaVA、Qwen-VL、InternVL和InstructBLIP）上，AttWarp持续提升了准确率，增强了组合推理能力，并减少了幻觉现象，其表现优于四种在测试时处理原始图像的竞争基线方法。这些结果表明，注意力引导的扭曲能够优先处理与查询相关的信息，同时保留上下文，并且同一MLLM在接收此类扭曲输入时表现更佳。