3D Object Detection on ScanNet v2 (test)
74.5mAP@0.25CAGroup3D
Evaluation Results
| Method | Links | ||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CAGroup3DType=Offline2024.03 | 74.5 | — | 60.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCAF3DType=Offline2024.03 | 70.7 | — | 56 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCAF3D-SV+OursType=Online, Input=single-view RGB-D images, Adaptation=Memory-based Adapters2024.03 | 70.5 | — | 49.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CentralizedNumber of Clients (C)=20, Data Heterogeneity (alpha)=0.252026.04 | 60.3 | — | 35.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 86.7 | 73.2 | 86.2 | |
| CentralizedNumber of Clients (C)=100, Data Heterogeneity (alpha)=0.12026.04 | 60.3 | — | 35.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 86.7 | 73.2 | 86.2 | |
| AnyViewType=Online2024.03 | 60.1 | — | 36 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| 3DGeoDetTrain/Test Views=20/50, 3D Supervision=Depth maps, Venue=TMM 20252025.12 | 59.6 | — | 34.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GVSynergy-DetTrain/Test Views=20/50, 3D Supervision=None2025.12 | 56.3 | — | 32.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ImGeoNetTrain/Test Views=20/50, 3D Supervision=Point clouds, Venue=ICCV 20232025.12 | 54.5 | — | 28.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoteNetAnnotation=Bounding Box (BBox)2022.03 | 54.2 | 110 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MVSDetTrain/Test Views=20/50, 3D Supervision=None, Venue=NeurIPS 20242025.12 | 54 | — | 29 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| NeRF-Det++Train/Test Views=50/50, 3D Supervision=None, Venue=TIP 20252025.12 | 53.9 | — | 29.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CN-RMATrain/Test Views=20/50, 3D Supervision=TSDF volume, Venue=CVPR 20242025.12 | 51 | — | 31 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| NeRF-DetTrain/Test Views=20/50, 3D Supervision=None, Venue=ICCV 20232025.12 | 50.4 | — | 25.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Fed3DNumber of Clients (C)=100, Data Heterogeneity (alpha)=0.12026.04 | 47.5 | — | 30.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.4 | 55.7 | 43.6 | |
| Back to Reality (BR)Annotation=Position-level (P-L)2022.03 | 47 | 5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VGGT-Detconfig=Basic backbone+AG+QD2026.03 | 46.9 | — | — | 27.8 | 82 | 61.9 | 74 | 50.3 | 18.3 | 17.6 | 36.9 | 2.6 | 40.3 | 66.9 | 31.4 | 49 | 32.9 | 90.3 | 50.8 | 83.2 | 27.6 | — | — | — | |
| FedDynNumber of Clients (C)=100, Data Heterogeneity (alpha)=0.12026.04 | 46.7 | — | 30 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.6 | 55 | 86.2 | |
| MOONNumber of Clients (C)=100, Data Heterogeneity (alpha)=0.12026.04 | 46.3 | — | 29.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83 | 54.8 | 86.2 | |
| Fed3DNumber of Clients (C)=20, Data Heterogeneity (alpha)=0.252026.04 | 46.1 | — | 26.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83 | 53 | 43.6 | |
| FedProxNumber of Clients (C)=100, Data Heterogeneity (alpha)=0.12026.04 | 45.8 | — | 28.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.3 | 53.8 | 86.2 | |
| SCAFFOLDNumber of Clients (C)=100, Data Heterogeneity (alpha)=0.12026.04 | 45.2 | — | 29.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 80.8 | 55.3 | 86.2 | |
| MOONNumber of Clients (C)=20, Data Heterogeneity (alpha)=0.252026.04 | 45 | — | 26.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 82 | 52.2 | 86.2 | |
| VGGT-Detconfig=Basic backbone+AG2026.03 | 44.2 | — | — | 25.9 | 84.7 | 53.7 | 73.8 | 46.3 | 17.6 | 13.7 | 26.4 | 2.2 | 40.5 | 66.5 | 18.7 | 41.5 | 45.2 | 84.6 | 45.3 | 84 | 24.3 | — | — | — | |
| ImVoxelNetTrain/Test Views=20/50, 3D Supervision=None, Venue=WACV 20222025.12 | 43.4 | — | 19.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FedAvgNumber of Clients (C)=100, Data Heterogeneity (alpha)=0.12026.04 | 43.4 | — | 29 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81 | 52.9 | 86.2 | |
| FedDynNumber of Clients (C)=20, Data Heterogeneity (alpha)=0.252026.04 | 43.3 | — | 22.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.4 | 49 | 86.2 | |
| MVSDetSG-Free setting=true2026.03 | 42.5 | — | — | 28 | 80.6 | 55.5 | 65.3 | 43.2 | 25.2 | 9.9 | 29.4 | 1.1 | 43.4 | 63.8 | 20.4 | 39.2 | 36.7 | 89.7 | 44.2 | 68.4 | 20.2 | — | — | — | |
| FedProxNumber of Clients (C)=20, Data Heterogeneity (alpha)=0.252026.04 | 42.3 | — | 26.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.5 | 51.9 | 86.2 | |
| SCAFFOLDNumber of Clients (C)=20, Data Heterogeneity (alpha)=0.252026.04 | 42.2 | — | 24.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.6 | 48.9 | 86.2 | |
| FCAF3D-SVType=Online, Input=single-view RGB-D images2024.03 | 41.9 | — | 20.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VGGT-Detconfig=Basic backbone2026.03 | 41.4 | — | — | 20.1 | 85.5 | 48.4 | 78.5 | 37 | 13.2 | 9.4 | 22.1 | 0.8 | 40.4 | 68.3 | 19.3 | 35.5 | 41.4 | 89.5 | 39.7 | 76.8 | 20 | — | — | — | |
| FedAvgNumber of Clients (C)=20, Data Heterogeneity (alpha)=0.252026.04 | 41.3 | — | 25.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.6 | 52 | 86.2 | |
| NeRF-DetSG-Free setting=true2026.03 | 41.2 | — | — | 24.9 | 85.3 | 54.4 | 64.3 | 42.2 | 24.3 | 12.7 | 31 | 1 | 39.3 | 62.6 | 17.3 | 41.4 | 32.8 | 88.4 | 39.7 | 58.7 | 21 | — | — | — | |
| FCAF3DSG-Free setting=true2026.03 | 40.6 | — | — | 28.1 | 81 | 47.2 | 70.7 | 40.9 | 17.1 | 13.2 | 22 | 0.7 | 45.6 | 65.4 | 8.3 | 32.3 | 39.1 | 83.6 | 46.9 | 72.7 | 15.6 | — | — | — | |
| ImVoxelNetSG-Free setting=true2026.03 | 35.2 | — | — | 20.2 | 76.3 | 45.8 | 60.5 | 36 | 16.3 | 9.6 | 17.9 | 0.3 | 23.3 | 53.1 | 8.1 | 29.4 | 35 | 83.2 | 43.1 | 61.3 | 14.4 | — | — | — | |
| P-L BaselineAnnotation=Position-level (P-L)2022.03 | 32.4 | 5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Local-trainingNumber of Clients (C)=20, Data Heterogeneity (alpha)=0.252026.04 | 19.1 | — | 10.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 50.2 | 46.2 | 86.2 | |
| Local-trainingNumber of Clients (C)=100, Data Heterogeneity (alpha)=0.12026.04 | 19.1 | — | 10.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 50.2 | 46.2 | 86.2 | |
| 3D-SISInput=Geo + RGB2021.12 | — | — | 22.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| 4DContrast + VoteNetInput=Geo only2021.12 | — | — | 40 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CSC + VoteNetInput=Geo only2021.12 | — | — | 39.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DSSInput=Geo + RGB2021.12 | — | — | 6.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| F-PointNetInput=Geo + RGB2021.12 | — | — | 10.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GSPNInput=Geo + RGB2021.12 | — | — | 17.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PointContrast + VoteNetInput=Geo only2021.12 | — | — | 38 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RandomRooms + VoteNetInput=Geo only2021.12 | — | — | 36.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| S-L BaselineAnnotation=Scene-level (S-L)2022.03 | — | 1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Scratch + VoteNetInput=Geo only2021.12 | — | — | 34.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoteNetInput=Geo + Height2021.12 | — | — | 33.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |