3D Object Detection on Waymo Open Dataset Vehicle LEVEL_2 v1.2 (test)
70.543D mAP OverallM3DETR
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| M3DETRinput=single point cloud frame2021.04 | 70.54 | 89.43 | 70.19 | 45.57 | 69.98 | 89.01 | 69.54 | 44.62 | |
| PV-RCNN*reproducibility=reproduced results, input=single point cloud frame2021.04 | 67.99 | 89.18 | 69.39 | 42.8 | 67.46 | 88.75 | 68.7 | 41.95 | |
| RangeDet2021.04 | 67.12 | 87.53 | 67.99 | 44.4 | 66.73 | 87.08 | 67.58 | 44.01 | |
| Det3D2021.04 | 65.21 | 87.93 | 63.8 | 38.2 | 64.29 | 87.3 | 62.36 | 36.87 | |
| PointPillars2021.04 | 60.56 | 81.28 | 52.35 | 30.87 | 60.1 | 80.8 | 51.87 | 30.31 | |
| StarNet2021.04 | 54.9 | 81.3 | 49.5 | 23 | 54.5 | 80.8 | 49 | 22.7 |