Towards Safe Autonomous Driving: Capture Uncertainty in the Deep Neural Network For Lidar 3D Vehicle Detection

TL;DR

This paper introduces methods to explicitly model and capture both epistemic and aleatoric uncertainties in deep neural network-based 3D vehicle detection from Lidar data, enhancing safety and performance in autonomous driving.

Contribution

It presents practical probabilistic approaches to quantify uncertainties in 3D object detection, addressing a gap in existing deep learning models for autonomous vehicles.

Findings

Epistemic uncertainty correlates with detection accuracy.

Aleatoric uncertainty varies with vehicle distance and occlusion.

Modeling aleatoric uncertainty improves detection performance by 1-5%.

Abstract

To assure that an autonomous car is driving safely on public roads, its object detection module should not only work correctly, but show its prediction confidence as well. Previous object detectors driven by deep learning do not explicitly model uncertainties in the neural network. We tackle with this problem by presenting practical methods to capture uncertainties in a 3D vehicle detector for Lidar point clouds. The proposed probabilistic detector represents reliable epistemic uncertainty and aleatoric uncertainty in classification and localization tasks. Experimental results show that the epistemic uncertainty is related to the detection accuracy, whereas the aleatoric uncertainty is influenced by vehicle distance and occlusion. The results also show that we can improve the detection performance by 1%-5% by modeling the aleatoric uncertainty.