🤖 AI Summary
Under adverse weather conditions, LiDAR fails while radar suffers from noise and artifacts, leading to poor cross-modal registration robustness and significantly lower localization accuracy compared to LiDAR–LiDAR systems. To address this, we propose a radar–LiDAR point cloud co-localization method. Our key contributions are: (1) the first open-source, differentiable ICP library enabling end-to-end geometric optimization; and (2) a scan-level, semantic-aware radar point weighting network that jointly performs data-driven denoising and geometric consistency enhancement. By synergistically integrating geometric constraints with semantic priors, our method substantially improves registration convergence speed and robustness on real-world autonomous driving datasets, reducing average localization error by 32.7%. The implementation is fully open-sourced to ensure reproducibility and facilitate further extension.
📝 Abstract
This paper presents a novel deep-learning-based approach to improve localizing radar measurements against lidar maps. This radar-lidar localization leverages the benefits of both sensors; radar is resilient against adverse weather, while lidar produces high-quality maps in clear conditions. However, owing in part to the unique artefacts present in radar measurements, radar-lidar localization has struggled to achieve comparable performance to lidar-lidar systems, preventing it from being viable for autonomous driving. This work builds on ICP-based radar-lidar localization by including a learned preprocessing step that weights radar points based on high-level scan information. To train the weight-generating network, we present a novel, stand-alone, open-source differentiable ICP library. The learned weights facilitate ICP by filtering out harmful radar points related to artefacts, noise, and even vehicles on the road. Combining an analytical approach with a learned weight reduces overall localization errors and improves convergence in radar-lidar ICP results run on real-world autonomous driving data. Our code base is publicly available to facilitate reproducibility and extensions.
