3D Object Detection on KITTI tracking Easy (val)
89.56AP (Car)3D-FCT
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| 3D-FCTbackbone=PV-RCNN2021.10 | 89.56 | 55.52 | 89.15 | 78.08 | |
| PV-RCNN + T losstracking loss=true2021.10 | 86.11 | 49.05 | 83.66 | 72.94 | |
| SECOND2021.10 | 85.33 | 46.73 | 78.89 | 70.32 | |
| PV-RCNN2021.10 | 84.54 | 48.99 | 82.63 | 72.05 | |
| PointRCNN2021.10 | 83.75 | 45.14 | 55.3 | 61.4 | |
| PointPillar2021.10 | 75.28 | 22.08 | 68.27 | 55.21 |