Relative Energy Learning for LiDAR out-of-distribution detection

LiDAR perception is central to GeoAI applications such as autonomous driving. However, current LiDAR perception models are trained under a closed-set assumption and remain vulnerable to out-of-distribution (OOD) objects that are not represented in training datasets. OOD detection is essential for trustworthy and safe AI by identifying unexpected inputs that differ from the training distribution, as it enables distinguishing rare anomalies from common classes and avoiding overconfident predictions for unknown samples. In this paper, we propose Relative Energy Learning (REL), a simple yet effective framework for OOD detection in LiDAR point clouds. REL leverages the energy gap between positive (in-distribution) and negative logits as a relative scoring function, mitigating calibration issues in raw energy values and improving robustness across various scenes. To address the absence of OOD samples during training, we propose a lightweight data synthesis strategy called Point Raise, which perturbs existing point clouds to generate auxiliary anomalies without altering the inlier semantics. Evaluated on SemanticKITTI and the Spotting the Unexpected (STU) benchmark, REL consistently outperforms existing methods by a large margin. Our results highlight that modeling relative energy, combined with simple synthetic outliers, provides a principled and scalable solution for reliable OOD detection in open-world autonomous driving. 1