发表机构
I3A, Universidad de Zaragoza; Niantic Spatial; University of Oxford（萨拉戈萨大学I3A研究所; Niantic Spatial公司; 牛津大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
AutoCompass是一种基于弱标签的监督方法，通过利用原始GPS标签、GPS容差区域及SLAM/SfM获取的相对位姿，提升神经地图匹配器的视觉定位精度，在两类基准测试中均优于同类方法。
AI 中文摘要
神经地图匹配器可估计图像相对于2D地图的3自由度（3-DoF）位姿，这类模型基于大规模地理参考图像数据集训练，其位置与航向标签常含噪声，会影响训练出的模型。为解决该问题，我们提出AutoCompass，一种从不精准绝对位姿标签训练神经地图匹配器的监督方法。首先，我们证明航向标签并非必需：从原始GPS标签训练的模型可自动学习预测精准航向；其次，在原始GPS周围定义容差区域可提升位置精度；第三，若有可用的训练图像间相对位姿（通过SLAM或SfM获取），我们的监督方法会利用这些更精准的训练信号。在驾驶与自我中心基准测试中，AutoCompass的性能始终优于那些过度依赖绝对位姿标签训练的同类模型。
英文摘要
Neural map matchers estimate an image's 3-DoF pose relative to a 2D map. These models are trained on large-scale datasets of geo-referenced images, whose position and heading labels often contain noise that affects the trained models. To address this, we present AutoCompass, a supervision approach for training neural map matchers from inaccurate absolute pose labels. First, we show that heading labels are unnecessary: trained from raw GPS labels, models learn to predict accurate headings, automatically. Second, defining a tolerance region around raw GPS improves positional accuracy. Third, if available, our supervision uses relative poses between training images, obtained via SLAM or SfM, which provide a more accurate training signal. Across driving and egocentric benchmarks, AutoCompass consistently outperforms counterparts trained with the usual strong reliance on absolute pose labels.
CommentsECCV 2026
