# Radar Fields: Frequency-Space Neural Scene Representations for FMCW Radar Source: [Radar Fields: Frequency-Space Neural Scene Representations for FMCW Radar]() Collection index: [llms.txt]() Data updated: 2025-10-26T02:42:27.853Z Authors: David Borts, Erich Liang, Tim Br\\"odermann, Andrea Ramazzina, Stefanie Walz, Edoardo Palladin, Jipeng Sun, David Bruggemann, Christos Sakaridis, Luc Van Gool, Mario Bijelic, Felix Heide Published: 2024-05-07T20:44:48.000Z ## Abstract Neural fields have been broadly investigated as scene representations for the reproduction and novel generation of diverse outdoor scenes, including those autonomous vehicles and robots must handle. While successful approaches for RGB and LiDAR data exist, neural reconstruction methods for radar as a sensing modality have been largely unexplored. Operating at millimeter wavelengths, radar sensors are robust to scattering in fog and rain, and, as such, offer a complementary modality to active and passive optical sensing techniques. Moreover, existing radar sensors are highly cost-effective and deployed broadly in robots and vehicles that operate outdoors. We introduce Radar Fields - a neural scene reconstruction method designed for active radar imagers. Our approach unites an explicit, physics-informed sensor model with an implicit neural geometry and reflectance model to directly synthesize raw radar measurements and extract scene occupancy. The proposed method does not rely on volume rendering. Instead, we learn fields in Fourier frequency space, supervised with raw radar data. We validate the effectiveness of the method across diverse outdoor scenarios, including urban scenes with dense vehicles and infrastructure, and in harsh weather scenarios, where mm-wavelength sensing is especially favorable. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Radar Novel View Synthesis on RobotCar (Scene 1)](): Radar Novel View Synthesis; RobotCar (Scene 1); SSIM: 0.3372. - [Radar Novel View Synthesis on RobotCar (Scene 2)](): Radar Novel View Synthesis; RobotCar (Scene 2); SSIM: 29.97. - [Radar Novel View Synthesis on RobotCar (Scene 4)](): Radar Novel View Synthesis; RobotCar (Scene 4); SSIM: 28.29. - [Radar Novel View Synthesis on Boreas Sun scenario](): Radar Novel View Synthesis; Boreas Sun scenario; SSIM: 36.41. - [Radar Novel View Synthesis on Boreas Snow scenario](): Radar Novel View Synthesis; Boreas Snow scenario; SSIM: 48.23. - [Radar Novel View Synthesis on Boreas Rain scenario](): Radar Novel View Synthesis; Boreas Rain scenario; SSIM: 39.05. - [Radar Novel View Synthesis on Boreas Static scenario](): Radar Novel View Synthesis; Boreas Static scenario; SSIM: 43.31. - [Radar Novel View Synthesis on RobotCar (Scene 3)](): Radar Novel View Synthesis; RobotCar (Scene 3); SSIM: 34.98. - [Radar Novel View Synthesis on RadarSim Dataset (Temporal Split)](): Radar Novel View Synthesis; RadarSim Dataset (Temporal Split); SSIM: 0.771. ## Original sources and implementation [Original paper]() [PDF]()