Learning an Uncertainty-Aware Object Detector for Autonomous Driving

TL;DR

This paper introduces an uncertainty-aware object detection method for autonomous driving that accounts for sensor noise and data incompleteness, enhancing detection accuracy and reliability.

Contribution

It proposes a novel approach to better learn probability distributions for object detection by considering noise in ground-truth labels, improving both uncertainty estimation and detection performance.

Findings

Improved detection accuracy over baseline methods

More reliable uncertainty estimates for autonomous systems

Enhanced robustness to sensor noise

Abstract

The capability to detect objects is a core part of autonomous driving. Due to sensor noise and incomplete data, perfectly detecting and localizing every object is infeasible. Therefore, it is important for a detector to provide the amount of uncertainty in each prediction. Providing the autonomous system with reliable uncertainties enables the vehicle to react differently based on the level of uncertainty. Previous work has estimated the uncertainty in a detection by predicting a probability distribution over object bounding boxes. In this work, we propose a method to improve the ability to learn the probability distribution by considering the potential noise in the ground-truth labeled data. Our proposed approach improves not only the accuracy of the learned distribution but also the object detection performance.