PAUL：不确定性引导的分区与增强用于噪声对应下的鲁棒跨视角地理定位
PAUL: Uncertainty-Guided Partition and Augmentation for Robust Cross-View Geo-Localization under Noisy Correspondence
针对跨视角地理定位中因GPS漂移导致的图像对空间错位（噪声对应）问题，提出基于数据不确定性和损失模式划分干净/噪声样本、并对高置信区域做局部增强的鲁棒训练框架，可复用于其他存在标注噪声或弱对应的视觉检索/匹配任务。
原文依据 this work presents the first systematic investigation of the Noisy Correspondence in Cross-View Geo-Localization (NC-CVGL) problem
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论跨视角地理定位是无人机导航、事件检测和航空测绘中的关键任务，旨在建立无人机拍摄图像与卫星图像之间的对应关系。现有大多数方法将跨视角数据嵌入联合特征空间，以最大化配对图像之间的相似度。然而，这些方法通常假设训练数据中的图像对完全对齐，这一假设在实际场景中很少成立。在真实条件下，城市峡谷效应、电磁干扰和恶劣天气等因素经常引起GPS漂移，导致系统性的对齐偏移，使得图像对之间仅存在部分对应关系。尽管这种噪声对应问题普遍存在，但当前研究对其关注有限。据我们所知，本文首次对跨视角地理定位中的噪声对应（NC-CVGL）问题进行了系统研究，专门针对由于GPS不准确导致相当一部分训练对存在空间错位的实际场景。为此，我们提出了PAUL（通过不确定性学习进行分区与增强），该框架通过三种协同机制实现噪声鲁棒学习：协同分区利用数据不确定性和损失模式将噪声样本与干净样本分离；协同增强通过局部评估增强高置信区域；协同训练通过双网络之间的相互监督来优化学习。与传统的过滤或重新标记噪声样本的噪声处理方法不同，PAUL通过这种三重协同机制有效利用了噪声数据。全面的实验验证了PAUL中各个组件的有效性，并在各种噪声比例下始终优于其他具有竞争力的噪声对应驱动方法。
英文原文摘要 1943 字符
Cross-view geo-localization is a critical task for UAV navigation, event detection, and aerial surveying, which establish correspondence between drone-captured and satellite imagery. Most existing approaches embed cross-view data into a joint feature space to maximize similarity between paired images. However, these methods typically assume perfect alignment of image pairs in training data, an assumption that rarely holds in practical scenarios. In real-world conditions, factors such as urban canyon effects, electromagnetic interference, and adverse weather frequently induce GPS drift, resulting in systematic **alignment shifts** where only partial correspondences exist between image pairs. Despite its prevalence, this source of noisy correspondence has received limited attention in current research.To our best knowledge, this work presents the first systematic investigation of the **Noisy Correspondence in Cross-View Geo-Localization (NC-CVGL)** problem, specifically addressing the practical scenario where a significant portion of training pairs exhibit spatial misalignment due to GPS inaccuracies. To this end, we propose **PAUL** (**P**artition and **A**ugmentation by **U**ncertainty **L**earning), a framework that achieves noise-robust learning through three coordinated mechanisms: **Co-partition** separates noisy from clean samples using data uncertainty and loss patterns; **Co-augmentation** enhances high-confidence regions via local assessment; and **Co-training** refines learning through mutual supervision between dual networks.Unlike conventional noise-handling methods that filter or relabel noisy samples, PAUL effectively utilizes noisy data through this triple collaborative mechanism. Comprehensive experiments validate the effectiveness of individual components in PAUL, which consistently achieves superior performance over other competitive noisy-correspondence-driven methods in various noise ratios.