3D Scene Generation on Replica (val)
0.0486FIDSemantic Generation
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Semantic Generationgeneration_mode=latent semantic only2025.09 | 0.0486 | 0.1193 | 0.0668 | |
| 2-stage Generationgeneration_mode=2-stage2025.09 | 0.8671 | 0.5245 | 0.2837 | |
| SceneSensetype=Baseline, representation=triplane- or voxel-based2025.09 | 31.095 | 89.7884 | 47.8399 | |
| SemCitytype=Baseline, representation=triplane- or voxel-based2025.09 | 50.2789 | 187.254 | 99.5552 |