Cross-Modal Masked Autoencoders for 4D Radar: Learning From Lidar to Boost Radar Object Detection Performance
IEEE Access · 2026 · 14 · 99413-99425
DOI: 10.1109/access.2026.3706123 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Patrick Palmer und 2 weiteren Autor:innen. Die Arbeit nennt 78 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Deutsche Forschungsgemeinschaft and Technische Universität Dortmund/TU Dortmund University within the funding program Open Access Costs.
Abstract (Originalsprache)
Radar-only object detection is experiencing growing interest due to the unique advantages of radar sensors, combined with recent advancements in improving their resolution. However, 3D object detection performance on radar remains limited compared to other sensor modalities, primarily due to the persisting sparsity and noise of radar point clouds. While several works have explored the use of different sensor modalities, such as lidar, to support the training of radar-only object detectors, they rely on labeled datasets, which limits the amount of usable data. To address this limitation, we draw inspiration from masked autoencoders and propose the radar-lidar autoencoder (RaLi-AE), a cross-modal autoencoder designed for self-supervised pre-training of radar-only object detectors. The main pre-training objective is the reconstruction of the dense lidar point cloud from the sparse radar point cloud. Our approach demonstrates an overall improvement of up to 3.1% points mean average precision on two different baseline methods and three datasets. Furthermore, we show that our method enables effective transfer learning across datasets.
Förderung
- Deutsche Forschungsgemeinschaft and Technische Universität Dortmund/TU Dortmund University within the funding program Open Access Costs
Zitieren
APA
Palmer, P., Krüger, M., & Bertram, T. (2026). Cross-Modal Masked Autoencoders for 4D Radar: Learning From Lidar to Boost Radar Object Detection Performance. IEEE Access, 14, 99413-99425. https://doi.org/10.1109/access.2026.3706123
Vancouver
Palmer P, Krüger M, Bertram T. Cross-Modal Masked Autoencoders for 4D Radar: Learning From Lidar to Boost Radar Object Detection Performance. IEEE Access. 2026;14:99413-99425. doi:10.1109/access.2026.3706123
BibTeX
@article{Palmer2026,
author = {Palmer, Patrick and Krüger, Martin and Bertram, Torsten},
title = {Cross-Modal Masked Autoencoders for 4D Radar: Learning From Lidar to Boost Radar Object Detection Performance},
journal = {IEEE Access},
year = {2026},
volume = {14},
pages = {99413--99425},
doi = {10.1109/access.2026.3706123}
}
EUVIMED