3D Multi-Object Tracking on nuScenes (val)
77.1AMOTAFusion-Poly
Evaluation results
| Method | Links | ||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Fusion-PolyDetector=CP & CR, Input=C + L2026.03 | 77.1 | — | — | 67.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 416 | — | — | — | — | — | — | — | — | — | — | |
| Fusion-PolyDetector=CP & CR, Input=C + L2026.03 | 77.1 | — | — | 67.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 416 | — | — | — | — | — | — | — | — | — | — | |
| CAMO-MOTDetector=BEVF & FC, Input=C + L2026.03 | 76.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 239 | — | — | — | — | — | — | — | — | — | — | |
| CAMO-MOTDetector=BEVF & FC, Input=C + L2026.03 | 76.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 239 | — | — | — | — | — | — | — | — | — | — | |
| Fast-PolyDevice=7945HX, Detector=LargeKernel3D2024.03 | 76 | — | — | 65.8 | — | — | — | — | — | — | — | — | — | — | — | — | 307 | — | 34.2 | — | — | — | — | — | — | — | — | — | — | — | |
| Easy-Poly (Ours)Detector=LargeKernel3D [6], Input Data=2D + 3D, FPS=34.92025.02 | 75.6 | — | 53.6 | 65.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 242 | — | — | — | — | — | — | — | — | — | — | |
| Poly-MOTDevice=7945HX, Detector=LargeKernel3D2024.03 | 75.2 | — | — | 54.1 | — | — | — | — | — | — | — | — | — | — | — | — | 252 | — | 8.6 | — | — | — | — | — | — | — | — | — | — | — | |
| Poly-MOTDetector=LargeKernel3D-L [6], Input Data=3D, FPS=8.62025.02 | 75.2 | — | 54.1 | 54.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 354 | — | — | — | — | — | — | — | — | — | — | |
| Easy-Poly (Ours)Detector=CenterPoint [32], Input Data=2D + 3D, FPS=34.62025.02 | 74.5 | — | 54.9 | 64.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 272 | — | — | — | — | — | — | — | — | — | — | |
| EMMS-MOTDetector=CP & CR, Input=C + L2026.03 | 74.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 247 | — | — | — | — | — | — | — | — | — | — | |
| EMMS-MOTDetector=CP & CR, Input=C + L2026.03 | 74.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 247 | — | — | — | — | — | — | — | — | — | — | |
| MC-TrackDetector=CP, Input=L2026.03 | 74 | — | — | 64 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 275 | — | — | — | — | — | — | — | — | — | — | |
| MC-TrackDetector=CP, Input=L2026.03 | 74 | — | — | 64 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 275 | — | — | — | — | — | — | — | — | — | — | |
| MCTrackFrame Approach=Global, Detection Source=CenterPoint [15]2026.06 | 74 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 275 | — | — | — | 85,900 | 13,083 | — | — | — | — | — | |
| CANMOTFrame Approach=Local, Joint Optimization (R and Q)=true2026.06 | 73.8 | — | 52.2 | — | 86.5 | — | — | 83.3 | — | — | — | 56.7 | 87.5 | 78.2 | 53 | 71.1 | — | — | — | 208 | — | — | 285 | 84,340 | 12,957 | 30 | 21 | 0.28 | 27 | 2,092.3 | |
| IMM-MOTFrame Approach=Global, Detection Source=CenterPoint [15]2026.06 | 73.8 | — | 51.6 | — | 86.4 | — | — | 82.8 | — | — | — | 56.4 | 87.3 | 78.9 | 51.6 | 71.6 | — | — | — | 326 | — | — | — | 85,913 | 13,433 | — | — | — | — | — | |
| Fast-PolyDevice=7945HX, Detector=CenterPoint2024.03 | 73.7 | — | — | 63.2 | — | — | — | — | — | — | — | — | — | — | — | — | 414 | — | 28.9 | — | — | — | — | — | — | — | — | — | — | — | |
| Fast-PolyDetector=CenterPoint [32], Input Data=3D, FPS=28.92025.02 | 73.7 | — | — | 63.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 414 | — | — | — | — | — | — | — | — | — | — | |
| Fast-PolyDetector=CP, Input=L2026.03 | 73.7 | — | — | 63.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 414 | — | — | — | — | — | — | — | — | — | — | |
| Fast-PolyDetector=CP, Input=L2026.03 | 73.7 | — | — | 63.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 414 | — | — | — | — | — | — | — | — | — | — | |
| CANMOTFrame Approach=Local2026.06 | 73.5 | — | 51.2 | — | 86.5 | — | — | 83.4 | — | — | — | 56.7 | 87.6 | 78.1 | 50.9 | 71.3 | — | — | — | 190 | — | — | 285 | 83,876 | 12,251 | 29 | 20 | 0.26 | 28 | 20.4 | |
| sharedFrame Approach=Local, Joint Optimization (R and Q)=true2026.06 | 73.3 | — | 52.2 | — | 86.2 | — | — | 83.1 | — | — | — | 56.2 | 86.9 | 77.7 | 51.6 | 71.3 | — | — | — | 226 | — | — | 316 | 84,550 | 13,413 | 30 | 21 | 0.26 | 25 | 1,902.5 | |
| Poly-MOTDevice=7945HX, Detector=CenterPoint2024.03 | 73.1 | — | — | 61.9 | — | — | — | — | — | — | — | — | — | — | — | — | 232 | — | 5.6 | — | — | — | — | — | — | — | — | — | — | — | |
| Poly-MOTDetector=CenterPoint [32], Input Data=3D, FPS=5.62025.02 | 73.1 | — | 52.1 | 61.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 281 | — | — | — | — | — | — | — | — | — | — | |
| DINO-MOTDetector=CP & DINO, Input=C + L2026.03 | 73.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 216 | — | — | — | — | — | — | — | — | — | — | |
| Poly-MOTDetector=CP, Input=L2026.03 | 73.1 | — | — | 61.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 232 | — | — | — | — | — | — | — | — | — | — | |
| DINO-MOTDetector=CP & DINO, Input=C + L2026.03 | 73.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 216 | — | — | — | — | — | — | — | — | — | — | |
| Poly-MOTDetector=CP, Input=L2026.03 | 73.1 | — | — | 61.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 232 | — | — | — | — | — | — | — | — | — | — | |
| sharedFrame Approach=Global, Joint Optimization (R and Q)=true2026.06 | 73 | — | 50.9 | — | 86.3 | — | — | 83 | — | — | — | 55.4 | 87.4 | 76.8 | 51.2 | 71 | — | — | — | 204 | — | — | 289 | 83,474 | 12,442 | 29 | 20 | 0.27 | 26 | 7,545.1 | |
| ShaSTADevice=A100, Detector=CenterPoint2024.03 | 72.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 10 | — | — | — | — | — | — | — | — | — | — | — | |
| Poly-MOTFrame Approach=Global, Detection Source=CenterPoint [15], Reproduced Status=true2026.06 | 72.8 | — | 51.7 | — | 86.3 | — | — | 83 | — | — | — | 54.4 | 87.3 | 76.8 | 51 | 71.2 | — | — | — | 235 | — | — | 295 | 84,094 | 13,062 | 29 | 21 | 1.25 | 25 | 6 | |
| ByteTrackv2Detector=CenterPoint2024.03 | 72.4 | — | — | 62.4 | — | — | — | — | — | — | — | — | — | — | — | — | 183 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OursModalities=LiDAR + Camera2025.05 | 72.2 | — | — | — | 85.6 | — | — | 77.9 | — | — | — | 6,420 | 83 | 75.6 | 49.6 | 69.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SC-TFrame Approach=Global, Covariance Sampling Data=Training2026.06 | 72.1 | — | 54.7 | — | 86.1 | — | — | 83 | — | — | — | 54.9 | 84.9 | 76 | 49.6 | 70.5 | — | — | — | 252 | — | — | 263 | 83,971 | 13,030 | 33 | 20 | 0.27 | 29 | 51.8 | |
| SC-TVFrame Approach=Global, Covariance Sampling Data=Trainval2026.06 | 72.1 | — | 54.8 | — | 86.1 | — | — | 83.1 | — | — | — | 54.9 | 84.9 | 75.9 | 49.5 | 70.4 | — | — | — | 248 | — | — | 249 | 83,808 | 12,840 | 33 | 20 | 0.27 | 30 | 48.7 | |
| CBMOTDetector=CenterPoint [32] & CenterTrack [35], Input Data=2D + 3D, FPS=802025.02 | 72 | — | 48.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 479 | — | — | — | — | — | — | — | — | — | — | |
| SC-TFrame Approach=Local, Covariance Sampling Data=Training2026.06 | 72 | — | 55.8 | — | 86.2 | — | — | 82.8 | — | — | — | 53.7 | 84.6 | 76.6 | 49.8 | 70.4 | — | — | — | 246 | — | — | 273 | 84,547 | 13,563 | 33 | 20 | 0.26 | 28 | 52.3 | |
| SC-VFrame Approach=Local, Covariance Sampling Data=Val2026.06 | 72 | — | 55.4 | — | 86.2 | — | — | 82.9 | — | — | — | 53.8 | 84.7 | 76.9 | 49.2 | 70.1 | — | — | — | 221 | — | — | 269 | 83,857 | 12,959 | 33 | 20 | 0.26 | 29 | 43.3 | |
| SC-TVFrame Approach=Local, Covariance Sampling Data=Trainval2026.06 | 71.9 | — | 55.8 | — | 86.2 | — | — | 82.9 | — | — | — | 53.7 | 84.6 | 76.6 | 49.3 | 70.4 | — | — | — | 240 | — | — | 271 | 84,372 | 13,381 | 33 | 20 | 0.26 | 28 | 49.4 | |
| SC-VFrame Approach=Global, Covariance Sampling Data=Val2026.06 | 71.6 | — | 55.2 | — | 86.1 | — | — | 82.8 | — | — | — | 52.7 | 84.8 | 75.7 | 49 | 70.2 | — | — | — | 271 | — | — | 264 | 84,307 | 13,533 | 33 | 20 | 0.27 | 29 | 42.3 | |
| 3DMOTFormerDevice=2080Ti, Detector=CenterPoint2024.03 | 71.2 | — | — | 60.7 | — | — | — | — | — | — | — | — | — | — | — | — | 341 | — | 54.7 | — | — | — | — | — | — | — | — | — | — | — | |
| EagerMOTDetector=CenterPoint [32] & Cascade R-CNN [4], Input Data=2D + 3D, FPS=132025.02 | 71.2 | — | 56.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 899 | — | — | — | — | — | — | — | — | — | — | |
| EagerMOTDetector=CP & CR, Input=C + L2026.03 | 71.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 899 | — | — | — | — | — | — | — | — | — | — | |
| EagerMOTDetector=CP & CR, Input=C + L2026.03 | 71.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 899 | — | — | — | — | — | — | — | — | — | — | |
| Immortal Trackersensor=LiDAR2021.11 | 70.2 | — | — | 60.1 | — | — | — | — | — | — | — | — | — | — | — | — | 385 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoxelNeXtinput=LIDAR-only2023.03 | 70.2 | — | 64 | 61.5 | — | — | — | — | — | — | — | — | — | — | — | — | 729 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoxelNeXtModalities=LIDAR2025.05 | 70.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SimpleTrackFrequency=10Hz, Detector=CenterPoint [43]2021.11 | 69.6 | — | 54.7 | 60.2 | — | — | — | — | — | — | — | — | — | — | — | — | 405 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SimpleTrackinput=LIDAR-only2023.03 | 69.6 | — | 54.7 | 60.2 | — | — | — | — | — | — | — | — | — | — | — | — | 405 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SimpleTrackDevice=9940X, Detector=CenterPoint2024.03 | 69.6 | — | — | 60.2 | — | — | — | — | — | — | — | — | — | — | — | — | 405 | — | 0.5 | — | — | — | — | — | — | — | — | — | — | — | |
| SimpleTrackDetector=CenterPoint [32], Input Data=3D, FPS=0.52025.02 | 69.6 | — | 54.7 | 60.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 405 | — | — | — | — | — | — | — | — | — | — | |
| OGR3MOTDetector=CenterPoint [43]2021.11 | 69.3 | — | 62.7 | 60.2 | — | — | — | — | — | — | — | — | — | — | — | — | 262 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OGR3MOTsensor=LiDAR, peer-reviewed=false2021.11 | 69.3 | — | — | 60.2 | — | — | — | — | — | — | — | — | — | — | — | — | 262 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OGR3MOTinput=LIDAR-only2023.03 | 69.3 | — | 62.7 | 60.2 | — | — | — | — | — | — | — | — | — | — | — | — | 262 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OGR3MOTDevice=TITANXP, Detector=CenterPoint2024.03 | 69.3 | — | — | 60.2 | — | — | — | — | — | — | — | — | — | — | — | — | 262 | — | 12.3 | — | — | — | — | — | — | — | — | — | — | — | |
| OGR3MOTDetector=CenterPoint [32], Input Data=3D, FPS=12.32025.02 | 69.3 | — | 62.7 | 60.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 262 | — | — | — | — | — | — | — | — | — | — | |
| Our proposed methodModalities=2D + 3D, Input detection=CenterPoint [1]2020.12 | 68.7 | — | — | — | 84.3 | — | — | 76.6 | — | — | — | 4,900 | 82 | 70.2 | 53.4 | 65.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SimpleTrackFrequency=2Hz, Detector=CenterPoint [43]2021.11 | 68.7 | — | 57.3 | 59.2 | — | — | — | — | — | — | — | — | — | — | — | — | 519 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CenterPoint++sensor=LiDAR2021.11 | 68.7 | — | — | 59.2 | — | — | — | — | — | — | — | — | — | — | — | — | 519 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Chiu et al.Modalities=LiDAR + Camera2025.05 | 68.7 | — | — | — | 84.3 | — | — | 76.6 | — | — | — | 4,900 | 82 | 70.2 | 53.4 | 65.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Our proposed methodModalities=3D, Input detection=CenterPoint [1]2020.12 | 67.7 | — | — | — | 84.2 | — | — | 75.2 | — | — | — | 4,700 | 81.9 | 66.8 | 53.5 | 65.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CBMOTDetector=CenterPoint [43]2021.11 | 67.5 | — | 59.1 | 58.3 | — | — | — | — | — | — | — | — | — | — | — | — | 494 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CBMOTsensor=LiDAR2021.11 | 67.5 | — | — | 58.3 | — | — | — | — | — | — | — | — | — | — | — | — | 494 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CBMOTinput=LIDAR-only2023.03 | 67.5 | — | 59.1 | 58.3 | — | — | — | — | — | — | — | — | — | — | — | — | 494 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CenterPointDetector=CenterPoint [43]2021.11 | 66.5 | — | 56.7 | 56.2 | — | — | — | — | — | — | — | — | — | — | — | — | 562 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CenterPointsensor=LiDAR2021.11 | 66.5 | — | — | 56.2 | — | — | — | — | — | — | — | — | — | — | — | — | 562 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CenterPointinput=LIDAR-only2023.03 | 66.5 | — | 56.7 | 56.2 | — | — | — | — | — | — | — | — | — | — | — | — | 562 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CenterPointDetector=CenterPoint [32], Input Data=3D2025.02 | 66.5 | — | 56.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 562 | — | — | — | — | — | — | — | — | — | — | |
| CenterPointModalities=3D, Input detection=CenterPoint [1]2020.12 | 65.9 | — | — | — | 84.2 | — | — | 77.3 | — | — | — | 4,370 | 80.2 | 59.2 | 51.5 | 65.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CenterPointModalities=LIDAR2025.05 | 65.9 | — | — | — | 84.2 | — | — | 77.3 | — | — | — | 4,370 | 80.2 | 59.2 | 51.5 | 65.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MotiontrackDetector=Motiontrack, Input=C + L2026.03 | 64.1 | — | — | 59.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MotiontrackDetector=Motiontrack, Input=C + L2026.03 | 64.1 | — | — | 59.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Chiu et al.Implementation Source=Own implementation (*), Detector=CenterPoint [43]2021.11 | 62.4 | — | 65.5 | 54.2 | — | — | — | — | — | — | — | — | — | — | — | — | 1,098 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Chiu et al.Implementation Source=from OGR3MOT [44], Detector=CenterPoint [43]2021.11 | 61.7 | — | 98.4 | 53.3 | — | — | — | — | — | — | — | — | — | — | — | — | 680 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Probabilistic TrackingModalities=3D, Input detection=CenterPoint [1], Implementation Note=Authors' implementation using [2] and [1]2020.12 | 61.4 | — | — | — | 78 | — | — | 69.8 | — | — | — | 3,870 | 79.1 | 52.8 | 49.4 | 62.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AB3DMOTImplementation Source=Own implementation (*), Detector=CenterPoint [43]2021.11 | 59.8 | — | 77.1 | 53.7 | — | — | — | — | — | — | — | — | — | — | — | — | 1,570 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MPN-BaselineImplementation Source=from OGR3MOT [44], Detector=CenterPoint [43]2021.11 | 59.3 | — | 83.2 | 51.4 | — | — | — | — | — | — | — | — | — | — | — | — | 1,079 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MPN-Baselineinput=LIDAR-only2023.03 | 59.3 | — | 83.2 | 51.4 | — | — | — | — | — | — | — | — | — | — | — | — | 1,079 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AB3DMOTImplementation Source=from OGR3MOT [44], Detector=CenterPoint [43]2021.11 | 57.8 | — | 80.7 | 51.4 | — | — | — | — | — | — | — | — | — | — | — | — | 1,275 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AB3DMOTinput=LIDAR-only2023.03 | 57.8 | — | 80.7 | 51.4 | — | — | — | — | — | — | — | — | — | — | — | — | 1,275 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ProbabilisticTrackingModalities=3D, Input detection=MEGVII [15]2020.12 | 56.1 | — | — | — | 73.5 | — | — | 75.5 | — | — | — | 2,720 | 74.1 | 50.6 | 33.7 | 58 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DualViewMapDetImage Encoder=ResNet-50, Input Resolution=256 × 7042026.04 | 51.3 | — | 1.109 | 45.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | 59.5 | — | 473 | — | 58.3 | — | — | — | — | — | — | — | — | |
| BridgeAD-BBackbone=R1012025.03 | 51.2 | — | 108 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 544 | — | — | — | — | — | — | — | — | — | — | |
| SparseDrive-BBackbone=ResNet-101, Resolution=512x14082024.05 | 50.1 | — | 108.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | 632 | 60.1 | — | — | — | — | — | — | — | — | — | — | — | — | |
| ConsisDriveReal=true, Gen.=true2026.02 | 49.8 | — | 135 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 525 | — | — | — | — | — | — | — | — | — | — | |
| InstaDriveReal=true, Generated=true2026.02 | 49.6 | — | 137.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 532 | — | — | — | — | — | — | — | — | — | — | |
| InstaDriveReal=true, Gen.=true2026.02 | 49.6 | — | 137.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 532 | — | — | — | — | — | — | — | — | — | — | |
| Sparse4Dv3Image Encoder=ResNet-50, Input Resolution=256 × 704, Source=Reported in the original paper2026.04 | 49 | — | 1.164 | 43.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | 57.4 | — | 430 | — | 66 | — | — | — | — | — | — | — | — | |
| PF-TrackResolution=Full (F)2023.02 | 47.9 | — | 1.227 | 43.5 | — | — | — | — | — | — | — | — | — | — | — | — | 181 | 59 | — | — | — | — | — | — | — | — | — | — | — | — | |
| UniAD-Stage1Encoder=RN101, Pre-train=ViDAR-3D-det2023.12 | 45.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DQTrack-PETRv2Encoder=V2-99, Pre-train=nus-3D-det [54]2023.12 | 44.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Sparse4Dv3Image Encoder=ResNet-50, Input Resolution=256 × 704, Source=Baseline trained with code provided by authors2026.04 | 44 | — | 1.184 | 40.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | 52.5 | — | 403 | — | 62.2 | — | — | — | — | — | — | — | — | |
| CC-3DTconcurrent_work=true2023.02 | 42.9 | — | 1.257 | 38.5 | — | — | — | — | — | — | — | — | — | — | — | — | 2,219 | 53.4 | — | — | — | — | — | — | — | — | — | — | — | — | |
| DORTImage Encoder=ResNet-50, Input Resolution=256 × 7042026.04 | 42.4 | — | 1.264 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 49.2 | — | — | — | — | — | — | — | — | — | — | — | — | |
| PF-TrackResolution=Small (S)2023.02 | 40.8 | — | 1.343 | 37.6 | — | — | — | — | — | — | — | — | — | — | — | — | 166 | 50.7 | — | — | — | — | — | — | — | — | — | — | — | — | |
| DQTrack-StereoEncoder=RN101, Pre-train=nus-3D-det [44]2023.12 | 40.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Fast-PolyDevice=7945HX, Detector=DETR3D2024.03 | 40.7 | — | — | 33.1 | — | — | — | — | — | — | — | — | — | — | — | — | 1,312 | — | 39.3 | — | — | — | — | — | — | — | — | — | — | — | |
| HiP-AD2025.03 | 40.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DiffusionDrive-HATStage=2nd2025.12 | 40.2 | — | — | 36.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SparseDrive-HATStage=1st2025.12 | 40 | — | — | 37.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |