Skip to content

Object-Centric LiDAR-to-Radar Distillation for 3D Object Detection

2026 · IEEE Signal Processing Letters · Vol 33, pp. 3586-3590 · 0 citations · 27 references

Abstract

Radar is an essential sensor for autonomous driving due to its cost-effectiveness and robustness under adverse weather conditions. However, 3D radar points are inherently sparse, noisy, and lack elevation information. As a result, they provide much weaker geometric cues than LiDAR measurements, making reliable representation learning fundamentally challenging. To address this issue, prior works have explored LiDAR-to-radar distillation, enriching radar representations with dense LiDAR geometric priors. However, the substantial modality gap between LiDAR and radar may make such fine-grained alignment unnecessarily restrictive and limit the effectiveness of knowledge transfer. In this letter, we present an object-centric distillation framework that emphasizes semantically meaningful object-level cues rather than dense voxel-level matching. Specifically, we perform object-wise distillation by aggregating teacher and student features within each object region using Gaussian-weighted pooling. We further introduce density-aware curriculum weighting, which adaptively modulates the distillation strength according to the radar point density of each object. Experimental results on the nuScenes dataset show that the proposed method surpasses RadarDistill by 32.1% in relative mAP, which validates the effectiveness of the proposed framework.

View source

Similar papers

Sep 2026

ESAFusion: LiDAR-4-D Radar Fusion via Local Geometric Complementation and Multiscale Adaptive Interaction for 3-D Object Detection

LiDAR--4-D radar fusion combines accurate spatial geometry with motion and reflectivity cues from radar, offering a promising solution for 3-D object detection in complex driving environments. However, sparse radar observations and differences in spatial sampling between the two modalities complicate reliable cross-mod...

Gang Ma, Sen-Jie Hu, Jun-Jie Liu et al. · 0 citations
Open access Sep 2026

TRaIL-Odom: Tightly Coupled Continuous Time Radar-IMU-LiDAR Odometry With Adaptive Doppler Weighting

Existing radar-LiDAR fusion methods rely on fixed residual weights, even though the informativeness of radar Doppler and LiDAR geometry is scan- and direction-dependent, leading to uniform radar weighting that misallocates Doppler information across translational directions. To address this limitation, we propose two d...

Chiyun Noh, T. Tuna, William Talbot et al. · 0 citations
Sep 2026

Foreground guided dual branch learning for sparse 4D radar 3D object detection.

Experimental results on VoD and TJ4DRadSet demonstrate that the proposed radar-only method achieves competitive performance, even comparable to several radar-camera fusion methods, while maintaining a favorable balance between detection accuracy and inference efficiency.

Yuan-Hang Wang, Yong-Hua Zhou, Yong-Nan Zhang et al. · 0 citations
#artificial intelligence Preprint Sep 2026

SARFusion: Scene-Aware Routing Fusion for Robust Camera-LiDAR 3D Object Detection

SARFusion is proposed, a robust 3D object detector that decouples object-query decoding into three parallel reasoning branches: a camera branch, a LiDAR branch, and a camera-LiDAR fusion branch that alleviates harmful cross-modal interference while preserving the benefits of multimodal fusion when complementary cues ar...

Yu-Ting Zhao, Zi-Yi Zheng, Shu-Xiao Li · 0 citations
Open access Aug 2026

SRCDet: sparse fusion of surround-view radar and camera for 3D object detection

A query-based multimodal fusion framework, termed SRCDet, is proposed for camera-4D radar fusion, which achieves consistent improvements across nearly all metrics and low error rates in clear and adverse weather conditions, highlighting its practical adaptability to automotive-grade systems and effectiveness in safety-...

Wen-Jin Ai, Lianqing Zheng, Long Yang et al. · 0 citations
Open access 2026

Feature Alignment for NeRF-Based 3-D Object Detection

Two lightweight and complementary modules to enhance voxel feature quality for NeRF-based 3D detection with consistent improvements over the NeRF-RPN baseline in both recall and precision are introduced.

Yu-Han Wang, Gang Liu · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.