3D Object Detection on View-of-Delft (VoD) Entire Annotated Area (val)
72.93mAP3DLXL
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| LXLModality=R+C, FPS=6.12023.07 | 72.93 | 72.18 | 58.3 | 88.31 | — | |
| FUTR3DModality=R+C, FPS=7.32023.07 | 69.32 | 78.66 | 43.1 | 86.19 | — | |
| RCFusionModality=R+C, FPS=N/A2023.07 | 69.23 | 71.87 | 47.5 | 88.33 | — | |
| BEVFusionModality=R+C, FPS=7.12023.07 | 68.52 | 70.21 | 45.86 | 89.48 | — | |
| LXL-RModality=R, FPS=44.72023.07 | 68.51 | 70.26 | 47.34 | 87.93 | — | |
| PointPillarsModality=R, FPS=N/A2023.07 | 67.48 | 70.15 | 47.22 | 85.07 | — | |
| R4DetModality=R+C, FPS=8.32026.03 | 66.69 | 66.9 | 55.42 | 77.75 | — | |
| PointPillarsModality=L2023.07 | 65.9 | 66.6 | 56.1 | 75.1 | — | |
| RadarPillarNetModality=R, FPS=N/A2023.07 | 65.86 | 71.65 | 42.8 | 83.14 | — | |
| CVFusionModality=R+C, FPS=5.42026.03 | 65.41 | 60.87 | 57.89 | 77.46 | — | |
| HyDRaInput=C+R2024.03 | 60.9 | 52.8 | 56.6 | 73.3 | — | |
| SDCMParams(MB)=39.11, FPS=14.22026.01 | 60.02 | — | — | — | — | |
| SDCMYear=-, Modality=R+C2026.01 | 60.02 | 55.28 | 53.75 | 71.04 | — | |
| SGDet3DYear=2025, Params(MB)=73.65, FPS=9.22026.01 | 59.75 | — | — | — | — | |
| SGDet3DYear=2025, Modality=R+C2026.01 | 59.75 | 53.16 | 49.98 | 76.11 | — | |
| SGDet3DModality=4D Radar + Camera, FPS=9.22026.02 | 59.75 | 53.16 | 49.98 | 76.11 | — | |
| SGDet3DModality=R+C, FPS=9.22026.03 | 59.75 | 53.16 | 49.98 | 76.11 | — | |
| HGSFusionModality=R+C2024.12 | 58.96 | 51.67 | 52.64 | 72.58 | — | |
| HGSFusionYear=2025, Params(MB)=64.58, FPS=10.32026.01 | 57.72 | — | — | — | — | |
| HSGFusionYear=2025, Modality=R+C2026.01 | 57.72 | 51.67 | 52.39 | 68.5 | — | |
| RCDFNetYear=2025, Params(MB)=N/A, FPS=12.92026.01 | 56.66 | — | — | — | — | |
| RCDFNetYear=2025, Modality=R+C2026.01 | 56.66 | 48.65 | 45.27 | 76.06 | — | |
| DSFusionYear=2025, Modality=R+C2026.01 | 56.37 | 45.75 | 50.71 | 72.64 | — | |
| LXLModality=R+C, FPS=6.12023.07 | 56.31 | 42.33 | 49.48 | 77.12 | — | |
| LXLModality=R+C2024.12 | 56.31 | 42.33 | 49.48 | 77.12 | — | |
| LXLYear=2023, Modality=R+C2026.01 | 56.31 | 42.33 | 49.48 | 77.12 | — | |
| LXLModality=4D Radar + Camera, FPS=6.12026.02 | 56.31 | 42.33 | 49.48 | 77.12 | — | |
| LXLModality=R+C, FPS=6.12026.03 | 56.31 | 42.33 | 49.48 | 77.12 | — | |
| LXLModality=R+C2023.07 | 56.3 | 42.3 | 49.5 | 77.1 | — | |
| LXLInput=C+R2024.03 | 56.3 | 42.3 | 49.5 | 77.1 | — | |
| SFGFusionYear=2025, Params(MB)=N/A, FPS=6.42026.01 | 55.76 | — | — | — | — | |
| SFGFusionYear=2025, Modality=R+C2026.01 | 55.76 | 48.3 | 43.6 | 75.54 | — | |
| MAFF-NetYear=2025, Modality=R2026.01 | 54.59 | 42.33 | 46.75 | 74.72 | — | |
| IS-FusionYear=2024, Params(MB)=48.37, FPS=7.22026.01 | 54.4 | — | — | — | — | |
| IS-FusionYear=2024, Modality=R+C2026.01 | 54.4 | 48.57 | 46.17 | 68.48 | — | |
| UniBEVFusionYear=2025, Modality=R+C2026.01 | 54.09 | 42.22 | 47.11 | 72.94 | — | |
| IMKDInput=C+R2025.12 | 53.81 | 47.55 | 45.51 | 68.4 | — | |
| DRIFT (attention)Pre-trained=true2026.03 | 53.1 | 41.2 | 43.6 | 74.5 | — | |
| PointAugmentingYear=2021, Modality=R+C2026.01 | 52.6 | 39.62 | 44.48 | 73.7 | — | |
| PointAugmentingModality=4D Radar + Camera, FPS=7.92026.02 | 52.6 | 39.62 | 44.48 | 73.7 | — | |
| DRIFT (attention)Pre-trained=false2026.03 | 52.6 | 41.4 | 42.2 | 74.3 | — | |
| SD4RModality=4D Radar, FPS=22.12026.02 | 51.81 | 41.04 | 43.41 | 70.98 | — | |
| Li et al.Year=2025, Modality=R+C2026.01 | 51.72 | 52.11 | 43.58 | 59.45 | — | |
| ZFusionYear=2025, Params(MB)=N/A, FPS=8.82026.01 | 51.14 | — | — | — | — | |
| ZFusionYear=2025, Modality=R+C2026.01 | 51.14 | 44.77 | 40.02 | 68.61 | — | |
| SMURFInput=R2024.03 | 51 | 42.3 | 39.1 | 71.5 | — | |
| SMURF2026.03 | 51 | 42.3 | 39.1 | 71.5 | — | |
| SMURFModality=4D Radar-only, Radar scans=5-scans, Input features=x, y, z, Doppler, RCS2023.07 | 50.97 | 42.31 | 39.09 | 71.5 | 56.77 | |
| SMURFYear=2023, Modality=R2026.01 | 50.97 | 42.31 | 39.09 | 71.5 | — | |
| SMURFModality=4D Radar, FPS=30.32026.02 | 50.97 | 42.31 | 39.09 | 71.5 | — | |
| SMURFModality=R2026.03 | 50.97 | 42.31 | 39.09 | 71.5 | — | |
| RadarPillars2026.03 | 50.7 | 41.1 | 38.6 | 72.6 | — | |
| Wang et al.Year=2025, Modality=R+C2026.01 | 50.58 | 42.24 | 40.38 | 69.12 | — | |
| MUFASAYear=2024, Modality=R2026.01 | 50.24 | 43.1 | 38.97 | 68.65 | — | |
| MUFASA2026.03 | 50.2 | 43.1 | 39 | 68.7 | — | |
| RCBEVDetModality=R+C2024.12 | 49.99 | 40.63 | 38.86 | 70.48 | — | |
| RCBEVDetYear=2024, Modality=R+C2026.01 | 49.99 | 40.63 | 38.86 | 70.48 | — | |
| RCBEVDetInput=C+R2025.12 | 49.99 | 40.63 | 38.86 | 70.48 | — | |
| RCBEVDetModality=4D Radar + Camera2026.02 | 49.99 | 40.63 | 38.86 | 70.48 | — | |
| RCBEVDetIn=C+R2026.04 | 49.99 | 40.63 | 38.86 | 70.48 | — | |
| RCFusionInput=C+R2024.03 | 49.7 | 41.7 | 39 | 68.3 | — | |
| RCFusionModality=Camera + 4D Radar fusion, Radar scans=5-scans2023.07 | 49.65 | 41.7 | 38.95 | 68.31 | — | |
| RCFusionModality=R+C, FPS=N/A2023.07 | 49.65 | 41.7 | 38.95 | 68.31 | — | |
| RCFusionModality=R+C2024.12 | 49.65 | 41.7 | 38.95 | 68.31 | — | |
| RCFusionYear=2023, Modality=R+C2026.01 | 49.65 | 41.7 | 38.95 | 68.31 | — | |
| RCFusionInput=C+R2025.12 | 49.65 | 41.7 | 38.95 | 68.31 | — | |
| RCFusionModality=4D Radar + Camera, FPS=9.02026.02 | 49.65 | 41.7 | 38.95 | 68.31 | — | |
| RCFusionModality=R+C, FPS=9.02026.03 | 49.65 | 41.7 | 38.95 | 68.31 | — | |
| RCFusionIn=C+R2026.04 | 49.65 | 41.7 | 38.95 | 68.31 | — | |
| TL-4DRCFModality=R+C2024.12 | 49.35 | 43.71 | 40.11 | 64.22 | — | |
| TL-4DRCFYear=2024, Modality=R+C2026.01 | 49.35 | 43.71 | 40.11 | 64.22 | — | |
| BEVFusionModality=R+C, FPS=7.12023.07 | 49.25 | 37.85 | 40.96 | 68.95 | — | |
| BEVFusionModality=R+C2024.12 | 49.25 | 37.85 | 40.96 | 68.95 | — | |
| BEVFusionYear=2023, Modality=R+C2026.01 | 49.25 | 37.85 | 40.96 | 68.95 | — | |
| BEVFusionModality=4D Radar + Camera, FPS=7.12026.02 | 49.25 | 37.85 | 40.96 | 68.95 | — | |
| BEVFusionModality=R+C, FPS=7.12026.03 | 49.25 | 37.85 | 40.96 | 68.95 | — | |
| FUTR3DModality=R+C, FPS=7.32023.07 | 49.03 | 46.01 | 35.11 | 65.98 | — | |
| FUTR3DModality=R+C2024.12 | 49.03 | 46.01 | 35.11 | 65.98 | — | |
| FUTR3DYear=2023, Modality=R+C2026.01 | 49.03 | 46.01 | 35.11 | 65.98 | — | |
| FUTR3DModality=4D Radar + Camera, FPS=11.02026.02 | 49.03 | 46.01 | 35.11 | 65.98 | — | |
| FUTR3DModality=R+C, FPS=7.32026.03 | 49.03 | 46.01 | 35.11 | 65.98 | — | |
| MMF-BEVIn=C+R2026.04 | 48.92 | 40.1 | 37.8 | 68.88 | — | |
| Duan et al.Year=2025, Modality=R2026.01 | 47.74 | 39.18 | 37.23 | 66.81 | — | |
| CenterPointPre-trained=true, re-implemented=true2026.03 | 47.7 | 39.7 | 35.5 | 67.9 | — | |
| CenterPointre-implemented=true2026.03 | 47.4 | 39.9 | 37.8 | 64.7 | — | |
| LXL-RModality=R, FPS=44.72023.07 | 46.84 | 32.75 | 39.65 | 68.13 | — | |
| LXL-RModality=4D Radar, FPS=44.72026.02 | 46.84 | 32.75 | 39.65 | 68.13 | — | |
| LXL - RModality=R2023.07 | 46.8 | 32.8 | 39.6 | 68.1 | — | |
| Ruddat et al.Year=2024, Modality=R2026.01 | 46.2 | 36.7 | 36.8 | 65 | — | |
| RadarPillarNetModality=R, FPS=N/A2023.07 | 46.01 | 39.3 | 35.1 | 63.63 | — | |
| RadarPillarNetModality=R2024.12 | 46.01 | 39.3 | 35.1 | 63.63 | — | |
| RadarPillarNetYear=2023, Modality=R2026.01 | 46.01 | 39.3 | 35.1 | 63.63 | — | |
| RadarPillarNetInput=R2025.12 | 46.01 | 39.3 | 35.1 | 63.63 | — | |
| RadarPillarNetModality=4D Radar, FPS=54.62026.02 | 46.01 | 39.3 | 35.1 | 63.63 | — | |
| RadarPillarNetModality=R, FPS=98.82026.03 | 46.01 | 39.3 | 35.1 | 63.63 | — | |
| MMF-BEVIn=C+R, Fusion_Mode=Directly concatenating image and radar BEV features2026.04 | 45.45 | 39.65 | 32 | 64.71 | — | |
| CenterPointYear=2021, Modality=R2026.01 | 45.42 | 32.74 | 38 | 65.51 | — | |
| CenterPointModality=4D Radar, FPS=38.32026.02 | 45.42 | 32.74 | 38 | 65.51 | — | |
| CenterPointModality=R2026.03 | 45.42 | 32.74 | 38 | 65.51 | — | |
| CenterPointinherited from [6]=true2026.03 | 45.4 | 32.7 | 38 | 65.5 | — |