Robust Monocular Localization in Sparse HD Maps Leveraging Multi-Task Uncertainty Estimation

TL;DR

This paper introduces a monocular localization method for urban autonomous driving that uses multi-task uncertainty estimation and sparse HD maps to achieve robust and precise 6D vehicle positioning.

Contribution

It presents a novel uncertainty-aware perception module and differentiable cost maps for improved monocular localization in sparse urban maps.

Findings

Achieves robust 6D localization in challenging urban scenarios

Performs well with sparse HD maps containing only lane borders and traffic lights

Outperforms existing methods on Lyft 5 dataset

Abstract

Robust localization in dense urban scenarios using a low-cost sensor setup and sparse HD maps is highly relevant for the current advances in autonomous driving, but remains a challenging topic in research. We present a novel monocular localization approach based on a sliding-window pose graph that leverages predicted uncertainties for increased precision and robustness against challenging scenarios and per frame failures. To this end, we propose an efficient multi-task uncertainty-aware perception module, which covers semantic segmentation, as well as bounding box detection, to enable the localization of vehicles in sparse maps, containing only lane borders and traffic lights. Further, we design differentiable cost maps that are directly generated from the estimated uncertainties. This opens up the possibility to minimize the reprojection loss of amorphous map elements in an association free and uncertainty-aware manner. Extensive evaluation on the Lyft 5 dataset shows that, despite the sparsity of the map, our approach enables robust and accurate 6D localization in challenging urban scenarios