Relative Energy Learning for LiDAR Out-of-Distribution Detection

2025-11-10

AI Summary

To address the high false-positive rate and overconfidence of models in out-of-distribution (OOD) object detection from 3D LiDAR point clouds, this paper proposes the Relative Energy Learning (REL) framework. REL constructs a robust OOD scoring mechanism by modeling the energy difference between positive and negative logits. It introduces a lightweight Point Raise strategy to synthesize semantically preserved virtual OOD samples without requiring ground-truth OOD annotations. Furthermore, REL integrates energy-based modeling with input perturbation augmentation to enhance generalization under open-world scenarios. Evaluated on SemanticKITTI and STU benchmarks, REL significantly reduces false positives while achieving state-of-the-art performance, demonstrating both effectiveness and reliability in OOD detection for autonomous driving perception systems.

Technology Category

• Computer Vision: 3D Computer Vision
• Reasoning under Uncertainty: Relational Probabilistic Models
• Search and Optimization: Learning to Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Out-of-distribution (OOD) detection is a critical requirement for reliable autonomous driving, where safety depends on recognizing road obstacles and unexpected objects beyond the training distribution. Despite extensive research on OOD detection in 2D images, direct transfer to 3D LiDAR point clouds has been proven ineffective. Current LiDAR OOD methods struggle to distinguish rare anomalies from common classes, leading to high false-positive rates and overconfident errors in safety-critical settings. 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.

Problem

• Detecting out-of-distribution objects in LiDAR point clouds
• Reducing false-positive rates for rare anomalies in autonomous driving
• Addressing calibration issues in raw energy-based OOD detection methods

Innovation

• Relative Energy Learning for LiDAR OOD detection
• Energy gap between logits as scoring function
• Point Raise synthesizes anomalies from point clouds