以杆为中心的描述符用于鲁棒的机器人定位：使用小杆地标（SPL）数据集评估杆在距离（PaD）观测
Pole-centric Descriptors for Robust Robot Localization: Evaluation under Pole-at-Distance (PaD) Observations using the Small Pole Landmark (SPL) Dataset
AI总结：
本文提出以小杆地标数据集为核心的评估框架，通过对比学习方法在5-10米范围内提升稀疏几何的鲁棒性，为机器人定位提供实证支持。
摘要：
尽管杆状结构被广泛认可为长期机器人定位的稳定几何锚点，但其识别可靠性在典型的大型城市环境中常见的杆在距离（PaD）观测下显著下降。本文将重点从描述符设计转向对描述符鲁棒性的系统研究。我们的主要贡献是建立了以小杆地标（SPL）数据集为核心的专门评估框架。该数据集通过基于自动跟踪的关联管道构建，捕捉了相同物理地标在多视角、多距离下的观测，而无需人工标注。使用该框架，我们对对比学习（CL）和监督学习（SL）范式进行了比较分析。我们的发现表明，CL在稀疏几何中诱导出更鲁棒的特征空间，特别是在5-10米范围内实现了更优的检索性能。本工作为在具有挑战性的现实世界场景中评估地标独特性提供了实证基础和可扩展的方法。
英文摘要：
While pole-like structures are widely recognized as stable geometric anchors for long-term robot localization, their identification reliability degrades significantly under Pole-at-Distance (Pad) observations typical of large-scale urban environments. This paper shifts the focus from descriptor design to a systematic investigation of descriptor robustness. Our primary contribution is the establishment of a specialized evaluation framework centered on the Small Pole Landmark (SPL) dataset. This dataset is constructed via an automated tracking-based association pipeline that captures multi-view, multi-distance observations of the same physical landmarks without manual annotation. Using this framework, we present a comparative analysis of Contrastive Learning (CL) and Supervised Learning (SL) paradigms. Our findings reveal that CL induces a more robust feature space for sparse geometry, achieving superior retrieval performance particularly in the 5--10m range. This work provides an empirical foundation and a scalable methodology for evaluating landmark distinctiveness in challenging real-world scenarios.
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。
