发表机构
Xiamen University（厦门大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对无监督LiDAR点云配准中几何模糊导致性能崩溃的问题，提出CAESAR框架，利用训练时语义引导、双线索重匹配和语义预测蒸馏，在KITTI和nuScenes上达到最先进性能且推理零开销。
AI 中文摘要
大规模LiDAR点云的无监督配准仍然具有挑战性，因为户外场景固有的几何模糊性会降低伪标签质量并导致次优收敛，特别是对于稀疏、低分辨率的扫描，如nuScenes数据集中的扫描。我们发现，配准模型内在地编码了与配准精度强相关的语义感知，尽管没有显式的语义监督。然而，这种原生感知是脆弱的：无监督设置中几何模糊性产生的噪声监督会迅速侵蚀学习到的语义结构，导致性能崩溃。为此，我们提出了CAESAR，一个由现成的3D分割模型仅在训练期间引导的教师-学生框架。我们观察到，潜在的内部点匹配常常被埋在噪声特征空间中几个虚假邻居之下，这促使我们提出双线索引导的重新匹配，通过重新选择而非简单拒绝来恢复它们。在此基础上，一种仅训练时的语义-几何标签挖掘方法执行轻量级、批次特定的教师细化，并在语义引导下挖掘可靠的伪标签。我们进一步引入语义预测蒸馏，以在特征空间中巩固学生的语义感知。在KITTI和nuScenes上的大量实验展示了最先进的性能，在具有挑战性的nuScenes基准上取得了显著提升。关键在于，CAESAR在推理时零开销，且不需要配准数据上的语义标注。代码将发布。
英文摘要
Unsupervised registration of large-scale LiDAR point clouds remains challenging due to the geometric ambiguity inherent in outdoor scenes, which degrades pseudo-label quality and leads to suboptimal convergence, particularly for sparse, low-resolution scans such as those from nuScenes. We reveal that registration models intrinsically encode semantic awareness that strongly correlates with registration accuracy, albeit without explicit semantic supervision. However, this native awareness is fragile: noisy supervision arising from geometric ambiguity in unsupervised settings rapidly erodes the learned semantic structure, causing performance collapse. To this end, we propose CAESAR, a teacher-student framework guided by an off-the-shelf 3D segmentation model exclusively during training. We observe that potential inlier matches are often buried just beneath a few spurious neighbors in the noisy feature space, motivating Dual-Cue Guided Re-Matching to recover them through reselection rather than simply rejecting. Building on this, a train-only Semantic-Geometric Label Mining performs lightweight, batch-specific teacher refinement and mines reliable pseudo-labels under semantic guidance. We further introduce Semantic Predictive Distillation to consolidate the student's semantic awareness in the feature space. Extensive experiments on KITTI and nuScenes demonstrate state-of-the-art performance, with pronounced gains on the challenging nuScenes benchmark. Crucially, CAESAR incurs zero inference overhead and requires no semantic annotations on the registration data. Code will be released.
CommentsAccepted to ECCV 2026
