MuViT：多分辨率视觉变换器用于跨尺度学习的显微镜分析
MuViT: Multi-Resolution Vision Transformers for Learning Across Scales in Microscopy
AI总结：
MuViT 通过多分辨率视觉变换器在显微镜分析中实现跨尺度学习，提升多分辨率信息利用能力。
摘要：
现代显微镜 routinely 产生 gigapixel 图像，这些图像包含从精细细胞形态到更广泛的组织组织的多种空间尺度的结构。许多分析任务需要结合这些尺度，但大多数视觉模型仅在单一分辨率上操作或从单一视角衍生多尺度特征，限制了它们利用显微镜数据固有多分辨率性质的能力。我们介绍了 MuViT，一种变换器架构，旨在融合来自同一基础图像的真实多分辨率观察。MuViT 将所有补丁嵌入到共享的世界坐标系统中，并将旋转位置嵌入扩展到这些坐标，使注意力能够在单个编码器中整合广域上下文与高分辨率细节。在合成基准、肾组织病理学和高分辨率小鼠脑显微镜数据上，MuViT 都在强大的 ViT 和 CNN 基线模型上实现了持续改进。多分辨率 MAE 预训练进一步产生了尺度一致的表示，增强了下游任务。这些结果表明，显式世界坐标建模提供了一种简单而强大的机制，用于在大规模显微镜分析中利用多分辨率信息。
英文摘要：
Modern microscopy routinely produces gigapixel images that contain structures across multiple spatial scales, from fine cellular morphology to broader tissue organization. Many analysis tasks require combining these scales, yet most vision models operate at a single resolution or derive multi-scale features from one view, limiting their ability to exploit the inherently multi-resolution nature of microscopy data. We introduce MuViT, a transformer architecture built to fuse true multi-resolution observations from the same underlying image. MuViT embeds all patches into a shared world-coordinate system and extends rotary positional embeddings to these coordinates, enabling attention to integrate wide-field context with high-resolution detail within a single encoder. Across synthetic benchmarks, kidney histopathology, and high-resolution mouse-brain microscopy, MuViT delivers consistent improvements over strong ViT and CNN baselines. Multi-resolution MAE pretraining further produces scale-consistent representations that enhance downstream tasks. These results demonstrate that explicit world-coordinate modelling provides a simple yet powerful mechanism for leveraging multi-resolution information in large-scale microscopy analysis.
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。
