3D Object Detection on Waymo Open Dataset (WOD) (val) (Class-Specific AP by L1/L2 and Mixed IoU)
48.08Vehicle 3D AP@0.7 L1OWL
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| OWLReference=-, Learning Protocol=Unsupervised2025.12 | 48.08 | 41.51 | 29.66 | 24.68 | 32.27 | 31.14 | 36.91 | 32.44 | |
| CPDReference=CVPR2024, Learning Protocol=Unsupervised2025.12 | 37.4 | 32.13 | 16.31 | 13.22 | 5.06 | 4.87 | 19.59 | 16.74 | |
| OYSTERReference=CVPR2023, Learning Protocol=Unsupervised2025.12 | 14.66 | 14.1 | 0.18 | 0.14 | 0.33 | 0.32 | 5.06 | 4.85 | |
| MODESTReference=CVPR2022, Learning Protocol=Unsupervised2025.12 | 6.46 | 5.8 | 0.17 | 0.1 | 1.14 | 1.1 | 2.59 | 2.33 |