ZODS-RS -- 面向遥感的零训练目标检测与分割
ZODS-RS -- Zero-training Oriented Detection & Segmentation for Remote Sensing
AI总结：
提出一种无需训练的封闭式管道ZODS-RS，通过原型纯化、旋转尺度等变匹配和不确定性感知像素合并，统一了遥感图像的水平框检测与实例分割，在多个数据集上取得优异性能。
摘要：
遥感与无人机应用需要模型能够跨平台和视角泛化，而无需特定任务训练。然而，无训练管道在处理有向几何、尺度/旋转变化以及拥挤的港口或机场时常常失败，并且很少统一检测与分割。我们提出ZODS-RS，一种无训练、封闭式的管道，输出水平框（HBB）和实例掩码。基于DINOv3密集特征和SAM风格的提议，ZODS-RS链式包含：PP（通过Tyler协方差进行原型纯化）、R-SEM（使用可分离核和全局匈牙利分配的旋转尺度等变匹配）以及UAM（具有自适应先验和可选负原型的不确定性感知逐像素合并）。一个轻量级的CWLA融合多个DINOv3层。在FAIR1M（HBB）上，我们获得$\mathrm{mAP}_{0.50:0.95}=\mathbf{13.06}$和$\mathrm{AP}_S=\mathbf{2.93}$（船舶/飞机类别平均）；在xView（HBB）上，我们报告$\mathrm{mAP}=\mathbf{16.69}$。在我们的无人机数据集上，ZODS-RS实现了掩码$\mathrm{mIoU}=\mathbf{31.10}$，并在单张5090上将小目标AP相对于Grounded-SAM提升了$\mathbf{+30.70}$。这项工作为航空影像中的水平框检测加实例分割提供了统一的、无需训练的解决方案；提供了与DINOv3紧密耦合的PP/R-SEM/UAM的显式封闭形式公式；并在小目标和拥挤目标以及跨域迁移下展示了一致的增益，同时保持部署简单。
英文摘要：
Remote-sensing and UAV applications need models that generalize across platforms and viewpoints without task-specific training. Yet training-free pipelines often falter on oriented geometry, scale/rotation variation, and crowded ports or airfields, and rarely unify detection and segmentation. We introduce ZODS-RS, a training-free, closed-form pipeline that outputs horizontal boxes (HBB) and instance masks. Built on DINOv3 dense features and SAM-style proposals, ZODS-RS chains: PP (prototype purification via Tyler covariance), R-SEM (rotation-scale equivariant matching with separable kernels and global Hungarian assignment), and UAM (uncertainty-aware pixelwise merging with adaptive priors and optional negative prototypes). A lightweight CWLA fuses multiple DINOv3 layers. On FAIR1M (HBB) we obtain $\mathrm{mAP}_{0.50:0.95}=\mathbf{13.06}$ and $\mathrm{AP}_S=\mathbf{2.93}$ \emph{(class-averaged over ship/airplane)}; on xView (HBB) we report $\mathrm{mAP}=\mathbf{16.69}$. On our UAV dataset, ZODS-RS achieves mask $\mathrm{mIoU}=\mathbf{31.10}$ and improves small-object AP by $\mathbf{+30.70}$ over Grounded-SAM on a single 5090. This work offers a unified, \emph{no-training} solution for horizontal-box detection plus instance segmentation in aerial imagery; provides explicit closed-form formulations for PP/R-SEM/UAM tightly coupled with DINOv3; and demonstrates \emph{consistent} gains on small and crowded targets and under cross-domain shifts while keeping deployment simple.
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。
