Learning to Identify Out-of-Distribution Objects for 3D LiDAR Anomaly Segmentation

2026-04-26

AI Summary

This work addresses the challenge of anomaly segmentation for unknown or out-of-distribution objects in 3D LiDAR point clouds by proposing the first method that directly models the distribution of normal categories in 3D feature space, thereby enabling robust perception through explicit constraints on anomalous samples. The core contributions include an unsupervised deep learning–based anomaly detection framework that integrates 3D semantic segmentation architecture with feature distribution modeling, and the construction of the first hybrid benchmark dataset combining real-world and synthetic data to encompass diverse, complex scenes and multiple types of out-of-distribution objects. Experimental results demonstrate that the proposed approach achieves state-of-the-art performance on existing datasets and validates its effectiveness and generalization capability on the newly introduced benchmark.

Technology Category

• Computer Vision: 3D Computer Vision
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Understanding the surrounding environment is fundamental in autonomous driving and robotic perception. Distinguishing between known classes and previously unseen objects is crucial in real-world environments, as done in Anomaly Segmentation. However, research in the 3D field remains limited, with most existing approaches applying post-processing techniques from 2D vision. To cover this lack, we propose a new efficient approach that directly operates in the feature space, modeling the feature distribution of inlier classes to constrain anomalous samples. Moreover, the only publicly available 3D LiDAR anomaly segmentation dataset contains simple scenarios, with few anomaly instances, and exhibits a severe domain gap due to its sensor resolution. To bridge this gap, we introduce a set of mixed real-synthetic datasets for 3D LiDAR anomaly segmentation, built upon established semantic segmentation benchmarks, with multiple out-of-distribution objects and diverse, complex environments. Extensive experiments demonstrate that our approach achieves state-of-the-art and competitive results on the existing real-world dataset and the newly introduced mixed datasets, respectively, validating the effectiveness of our method and the utility of the proposed datasets. Code and datasets are available at https://simom0.github.io/lido-page/.

Problem

• 3D LiDAR
• anomaly segmentation
• out-of-distribution detection
• autonomous driving
• 3D perception

Innovation

• 3D LiDAR anomaly segmentation
• out-of-distribution detection
• feature space modeling
• mixed real-synthetic dataset
• autonomous driving perception