2D Object Detection on KITTI (val2)
97.2AP2DDeep3DBox
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Deep3DBoxIoU threshold=0.7, Criteria=Moderate, Category=Car2020.01 | 97.2 | 96.68 | — | — | — | |
| Ours(3D)IoU threshold=0.7, Criteria=Moderate, Category=Car, Method Variant=2D bounding box of the projected 3D box2020.01 | 92.08 | 89.4 | — | — | — | |
| Ours(2D)IoU threshold=0.7, Criteria=Moderate, Category=Car, Method Variant=keypoint detection network2020.01 | 91.85 | 89.22 | — | — | — | |
| DeepMANTAIoU threshold=0.7, Criteria=Moderate, Category=Car2020.01 | 91.01 | 90.66 | — | — | — | |
| GS3DIoU threshold=0.7, Criteria=Moderate, Category=Car2020.01 | 90.02 | 89.13 | — | — | — | |
| Deep3DBox2020.08 | — | — | 98.84 | 97.2 | 81.16 | |
| Deep3DBox + RAR-NetRefined by=RAR-Net2020.08 | — | — | 98.84 | 97.2 | 81.16 | |
| GS3D2020.08 | — | — | 81.66 | 71.68 | 62.8 | |
| GS3D + RAR-NetRefined by=RAR-Net2020.08 | — | — | 81.66 | 71.68 | 62.8 |