RadarGen: Automotive Radar Point Cloud Generation from Cameras

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Borreda, Tomer, Ding, Fangqiang, Fidler, Sanja, Huang, Shengyu, Litany, Or
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908723454148608
author Borreda, Tomer
Ding, Fangqiang
Fidler, Sanja
Huang, Shengyu
Litany, Or
author_facet Borreda, Tomer
Ding, Fangqiang
Fidler, Sanja
Huang, Shengyu
Litany, Or
contents We present RadarGen, a diffusion model for synthesizing realistic automotive radar point clouds from multi-view camera imagery. RadarGen adapts efficient image-latent diffusion to the radar domain by representing radar measurements in bird's-eye-view form that encodes spatial structure together with radar cross section (RCS) and Doppler attributes. A lightweight recovery step reconstructs point clouds from the generated maps. To better align generation with the visual scene, RadarGen incorporates BEV-aligned depth, semantic, and motion cues extracted from pretrained foundation models, which guide the stochastic generation process toward physically plausible radar patterns. Conditioning on images makes the approach broadly compatible, in principle, with existing visual datasets and simulation frameworks, offering a scalable direction for multimodal generative simulation. Evaluations on large-scale driving data show that RadarGen captures characteristic radar measurement distributions and reduces the gap to perception models trained on real data, marking a step toward unified generative simulation across sensing modalities.
format Preprint
id arxiv_https___arxiv_org_abs_2512_17897
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle RadarGen: Automotive Radar Point Cloud Generation from Cameras
Borreda, Tomer
Ding, Fangqiang
Fidler, Sanja
Huang, Shengyu
Litany, Or
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Robotics
We present RadarGen, a diffusion model for synthesizing realistic automotive radar point clouds from multi-view camera imagery. RadarGen adapts efficient image-latent diffusion to the radar domain by representing radar measurements in bird's-eye-view form that encodes spatial structure together with radar cross section (RCS) and Doppler attributes. A lightweight recovery step reconstructs point clouds from the generated maps. To better align generation with the visual scene, RadarGen incorporates BEV-aligned depth, semantic, and motion cues extracted from pretrained foundation models, which guide the stochastic generation process toward physically plausible radar patterns. Conditioning on images makes the approach broadly compatible, in principle, with existing visual datasets and simulation frameworks, offering a scalable direction for multimodal generative simulation. Evaluations on large-scale driving data show that RadarGen captures characteristic radar measurement distributions and reduces the gap to perception models trained on real data, marking a step toward unified generative simulation across sensing modalities.
title RadarGen: Automotive Radar Point Cloud Generation from Cameras
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Robotics
url https://arxiv.org/abs/2512.17897