Learning an Uncertainty-Aware Object Detector for Autonomous Driving
This work addresses uncertainty estimation for autonomous driving systems, which is crucial for safe vehicle reactions, but it appears incremental as it builds on prior methods for predicting bounding box distributions.
The paper tackles the problem of object detection uncertainty in autonomous driving by proposing a method that accounts for noise in ground-truth labels, resulting in improved accuracy of the learned probability distribution and enhanced object detection performance.
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.