← Back to Computer Vision cs.CV
Why radar might save autonomous cars in bad weather
Melih Yazgan, Iramm Hamdard, Qiyuan Wu, J. Marius Zoellner
May 29, 2026
Autonomous vehicles lose perception ability in fog and rain because cameras and lidar degrade. This work integrates 4D imaging radar—which sees through weather via radio waves and motion—into multi-agent perception systems with a Doppler-guided attention mechanism. Two new benchmarks (OPV2V-R, Adver-City-R) with simulated weather allow testing. Results show radar can replace or complement lidar, with major robustness gains in adverse conditions and validation on real truck data.
Read the original paper →