Text-to-3D Generation on 30 multi-object scenes
94.2CLIP R1-PrecisionGraphDreamer
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GraphDreamertraining protocol=two-stage2023.11 | 94.2 | 0.3326 | |
| MVDream2023.11 | 89.2 | 0.3102 | |
| Magic3Dtraining protocol=two-stage2023.11 | 87.5 | 0.3267 | |
| DreamFusiontraining steps=20K, training protocol=two-stage2023.11 | 80 | 0.2743 | |
| Fantasia3Dtraining steps=20K, initialization=default sphere (radius 0.5)2023.11 | 71.7 | 0.2243 |