HD-OOD3D: Supervised and Unsupervised Out-of-Distribution object detection in LiDAR data

TL;DR

HD-OOD3D introduces a two-stage LiDAR-based 3D object detection method capable of identifying unknown objects, with improved robustness through unsupervised learning and critical evaluation of existing protocols.

Contribution

The paper proposes a novel two-stage approach for 3D OOD detection in LiDAR data and explores unsupervised training strategies for unknown object detection.

Findings

Two-stage methods outperform single-stage in OOD detection.

Top-5 auto-labelling improves pseudo-label quality.

Evaluation protocol impacts hyperparameter tuning.

Abstract

Autonomous systems rely on accurate 3D object detection from LiDAR data, yet most detectors are limited to a predefined set of known classes, making them vulnerable to unexpected out-of-distribution (OOD) objects. In this work, we present HD-OOD3D, a novel two-stage method for detecting unknown objects. We demonstrate the superiority of two-stage approaches over single-stage methods, achieving more robust detection of unknown objects while addressing key challenges in the evaluation protocol. Furthermore, we conduct an in-depth analysis of the standard evaluation protocol for OOD detection, revealing the critical impact of hyperparameter choices. To address the challenge of scaling the learning of unknown objects, we explore unsupervised training strategies to generate pseudo-labels for unknowns. Among the different approaches evaluated, our experiments show that top-5 auto-labelling offers more promising performance compared to simple resizing techniques.