Simulate Any Radar: Attribute-Controllable Radar Simulation via Waveform Parameter Embedding

TL;DR

SA-Radar introduces a controllable radar simulation method that combines physics-based and generative approaches through waveform-parameterized attribute embedding, enabling realistic and versatile radar data generation for autonomous driving.

Contribution

It proposes a novel waveform-parameterized attribute embedding framework and ICFAR-Net for efficient, controllable radar simulation without detailed hardware specifications.

Findings

Generated radar data improves downstream detection and segmentation tasks.

SA-Radar enables simulation across diverse sensor viewpoints and scenes.

The approach outperforms traditional simulators in realism and efficiency.

Abstract

We present SA-Radar (Simulate Any Radar), a radar simulation approach that enables controllable and efficient generation of radar cubes conditioned on customizable radar attributes. Unlike prior generative or physics-based simulators, SA-Radar integrates both paradigms through a waveform-parameterized attribute embedding. We design ICFAR-Net, a 3D U-Net conditioned on radar attributes encoded via waveform parameters, which captures signal variations induced by different radar configurations. This formulation bypasses the need for detailed radar hardware specifications and allows efficient simulation of range-azimuth-Doppler (RAD) tensors across diverse sensor settings. We further construct a mixed real-simulated dataset with attribute annotations to robustly train the network. Extensive evaluations on multiple downstream tasks-including 2D/3D object detection and radar semantic segmentation-demonstrate that SA-Radar's simulated data is both realistic and effective, consistently improving model performance when used standalone or in combination with real data. Our framework also supports simulation in novel sensor viewpoints and edited scenes, showcasing its potential as a general-purpose radar data engine for autonomous driving applications. Code and additional materials are available at https://zhuxing0.github.io/projects/SA-Radar.