Out-of-Distribution Detection for LiDAR-based 3D Object Detection

TL;DR

This paper addresses the challenge of detecting out-of-distribution inputs in LiDAR-based 3D object detection for autonomous driving, proposing adapted detection methods and a new evaluation technique to improve safety.

Contribution

It formulates OOD detection for 3D object detection, adapts existing methods with a novel feature extraction approach, and introduces a technique to generate OOD objects for evaluation.

Findings

Different OOD detection methods have biases toward specific OOD objects

Combining multiple OOD detection methods improves detection robustness

Evaluation on KITTI dataset highlights the need for further research in this area

Abstract

3D object detection is an essential part of automated driving, and deep neural networks (DNNs) have achieved state-of-the-art performance for this task. However, deep models are notorious for assigning high confidence scores to out-of-distribution (OOD) inputs, that is, inputs that are not drawn from the training distribution. Detecting OOD inputs is challenging and essential for the safe deployment of models. OOD detection has been studied extensively for the classification task, but it has not received enough attention for the object detection task, specifically LiDAR-based 3D object detection. In this paper, we focus on the detection of OOD inputs for LiDAR-based 3D object detection. We formulate what OOD inputs mean for object detection and propose to adapt several OOD detection methods for object detection. We accomplish this by our proposed feature extraction method. To evaluate OOD detection methods, we develop a simple but effective technique of generating OOD objects for a given object detection model. Our evaluation based on the KITTI dataset shows that different OOD detection methods have biases toward detecting specific OOD objects. It emphasizes the importance of combined OOD detection methods and more research in this direction.