从面板到像素：面向生物医学科学文献的放大式视觉语言预训练
From Panel to Pixel: Zoom-In Vision-Language Pretraining from Biomedical Scientific Literature
把多面板科研图拆成 figure/panel/region 三级图文对，为视觉语言预训练提供细粒度、可复用的数据构造与多粒度目标统一思路，可直接迁移到任何含密集标注结构的图文语料（如说明书、海报、UI 截图）。
原文依据 constructs aligned image--text pairs at the figure, panel, and region levels, preserving local semantics
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论基于科学文献训练生物医学视觉语言模型正受到越来越多的关注。然而，多数流程将丰富的多面板图与长文本说明压缩为粗糙的图级图文对，丢弃了临床医生放大观察局部结构时所依赖的细粒度对应关系。我们提出 Panel2Patch，一个从多面板、标记密集的生物医学图及其周边文本中挖掘层次结构、并将其转化为多粒度监督的数据流程。给定图与说明文本，Panel2Patch 解析版面、面板与视觉标记，随后在图级、面板级与区域级构建对齐的图像—文本对，从而保留局部语义，而不是把每张图当作单个样本处理。基于该语料，我们提出一种粒度感知的预训练策略，在共享嵌入空间中将从粗粒度的教学式描述到细粒度的区域聚焦短语的异构目标统一起来。将 Panel2Patch 应用于一小部分文献图即可取得显著优于既有流程的性能，表明利用层次化的图结构能够以更少的预训练数据提供更有效的监督。
英文原文摘要 1211 字符
There is growing interest in biomedical vision--language models trained on scientific literature. However, most pipelines compress rich multi-panel figures and long captions into coarse figure-level pairs, discarding the fine-grained correspondences clinicians rely on when zooming into local structures. We introduce Panel2Patch, a data pipeline that mines hierarchical structure from multi-panel, marker-heavy biomedical figures and their surrounding text, and converts them into multi-granular supervision. Given figures and captions, Panel2Patch parses layouts, panels, and visual markers, then constructs aligned image--text pairs at the figure, panel, and region levels, preserving local semantics instead of treating each figure as a single sample. Built on this corpus, we develop a granularity-aware pretraining strategy that unifies heterogeneous objectives from coarse didactic descriptions to fine region-focused phrases in a shared embedding space. Applying Panel2Patch to a small subset of literature figures yields substantially better performance than prior pipelines, demonstrating that exploiting hierarchical figure structure can provide more effective supervision with less pretraining data.