3D Object Detection on OMNI3D
25.4AP (3D)GR3D-8B
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GR3D-8B2026.05 | 25.4 | — | |
| DetAny3DBase detector=Cube R-CNN2026.05 | 24.92 | — | |
| Cube R-CNNhandles focal length variance (f)=true, virtual camera (+vc)=true2022.07 | 23.3 | — | |
| Cube R-CNN2026.05 | 23.26 | — | |
| OVMono3DBase detector=Cube R-CNN2026.05 | 22.98 | — | |
| PGDhandles focal length variance (f)=true, virtual camera (+vc)=true2022.07 | 15.4 | — | |
| PGDhandles focal length variance (f)=false, virtual camera (+vc)=false2022.07 | 11.2 | — | |
| FCOS3Dhandles focal length variance (f)=true, virtual camera (+vc)=true2022.07 | 10.6 | — | |
| SMOKEhandles focal length variance (f)=true, virtual camera (+vc)=true2022.07 | 10.4 | — | |
| SMOKE2026.05 | 10.4 | — | |
| FCOS3Dhandles focal length variance (f)=false, virtual camera (+vc)=false2022.07 | 9.8 | — | |
| SMOKEhandles focal length variance (f)=false, virtual camera (+vc)=false2022.07 | 9.6 | — | |
| ImVoxelNethandles focal length variance (f)=true, virtual camera (+vc)=false2022.07 | 9.4 | — | |
| ImVoxelNet2026.05 | 9.4 | — | |
| 3DiffTectionBackbone=StableDiff+Geo-Ctr, Pre-training Data=LION5B & Aktscn, Pre-training Task=Generation, Trained Module=3D Head2023.11 | — | 17.21 | |
| 3DiffTectionBackbone=StableDiff+Geo-Ctr, Pre-training Data=LION5B & Aktscn, Pre-training Task=Generation, Trained Module=Sem-Ctr+3D Head2023.11 | — | 22.71 | |
| CubeRCNNBackbone=DLA34, Pre-training Data=ImageNet & Aktscn, Pre-training Task=Classification & 3D det., Trained Module=3D Head2023.11 | — | 16.13 | |
| DIFT-SDBackbone=StableDiff, Pre-training Data=LION5B (Schuhmann et al., 2022), Pre-training Task=Generation, Trained Module=3D Head2023.11 | — | 15.94 |