Research map / Detection and segmentation / 3D detection and driving
R4Det: 4D Radar-Camera Fusion for High-Performance 3D Object Detection
Zhongyu Xia, Yousen Tang, Yongtao Wang, Zhifeng Wang, Weijun Qin
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Panoramic Depth Fusion (ablation) | SGDet3D baseline · TJ4DRadSet | mAPBEV: 45.15 -> 46.86 | Improved |
| add dense metric loss (lambda_dense) to PDF (ablation) | PDF · TJ4DRadSet | mAPBEV: 45.15 -> 46.08 | Improved |
| add relative ranking loss to PDF (ablation) | PDF · TJ4DRadSet | mAPBEV: 46.08 -> 46.86 | Improved |
| add Deformable Gated Temporal Fusion (ablation) | SGDet3D + PDF · TJ4DRadSet | mAPBEV: 46.86 -> 50.41 | Improved |
| add ConvGRU gating on top of concat+DCN (ablation) | DGTF · TJ4DRadSet | mAP3D: 43.32 -> 44.86 | Improved |
| add Squeeze-and-Excitation to DGTF (ablation) | DGTF · TJ4DRadSet | mAP3D: 44.86 -> 42.47 | Hurt |
| add concat temporal fusion (ablation) | SGDet3D + PDF · TJ4DRadSet | mAPBEV: 46.86 -> 47.82 | Improved |
| add Instance-Guided Dynamic Refinement (ablation) | SGDet3D + PDF + DGTF · TJ4DRadSet | mAPBEV: 50.41 -> 54.07 | Improved |
| add foreground gating to naive prototype fusion (ablation) | IGDR · TJ4DRadSet | mAPBEV: 51.29 -> 51.60 | No clear effect |
| replace MLP generator with Conv calibration generator (ablation) | IGDR · TJ4DRadSet | mAPBEV: 52.48 -> 54.07 | Improved |
| replace attention generator with Conv generator (ablation) | IGDR · TJ4DRadSet | mAPBEV: 52.52 -> 54.07 | Improved |
| plug R4Det modules into BEVFusion | BEVFusion · VoD val (EAA) | mAP: 49.25 -> 55.59 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- CVFusion — CVFusion: Cross-View Fusion of 4D Radar and Camera for 3D Object Detection (ICCV 2025) · better on 5 of 5 shared results
- RCBEVDet — RCBEVDet: Radar-camera Fusion in Bird's Eye View for 3D Object Detection (CVPR 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- CRT-Fusion: Camera, Radar, Temporal Fusion Using Motion Information for 3D Object Detection (NeurIPS 2024 · similarity 0.88)
- RCBEVDet: Radar-camera Fusion in Bird's Eye View for 3D Object Detection (CVPR 2024 · similarity 0.87)
- L4DR: LiDAR-4DRadar Fusion for Weather-Robust 3D Object Detection (AAAI 2025 · similarity 0.87)
- RaCFormer: Towards High-Quality 3D Object Detection via Query-based Radar-Camera Fusion (CVPR 2025 · similarity 0.87)
- RobuRCDet: Enhancing Robustness of Radar-Camera Fusion in Bird's Eye View for 3D Object Detection (ICLR 2025 · similarity 0.87)
- TacoDepth: Towards Efficient Radar-Camera Depth Estimation with One-stage Fusion (CVPR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
