| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Multi-view generation | 3D-FUTURE | PSNR23.5001 | 9 | |
| Human-Object Interaction Generation | 3D-FUTURE (test) | Ts0 | 7 | |
| Classification | 3D-FUTURE (test) | Accuracy86.1 | 6 | |
| Furniture Synthesis | 3D-FUTURE 4 (unseen) | FID5.03 | 5 | |
| Compositional 3D Scene Generation | 3D-FUTURE first 200 scenes (test) | Chamfer Distance (CD)56,240 | 4 | |
| Human-Object Interaction Generation | 3D-FUTURE 9 (test) | Hfeet4.06 | 4 | |
| Human-Object Interaction Generation | 3D-FUTURE generalization (test) | FID0.154 | 3 | |
| Human-Object Interaction Generation | 3D-FUTURE (unseen objects, active frames) | Dhand6.05 | 2 | |
| Long-term Human-Object Interaction Synthesis | 3D-FUTURE | Temporal Start Error (Ts)2.41 | 2 |