| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| NeRF4 | DTC123 | CLIP-Similarity0.84 | 12 | 4d ago | |
| User Study (test) | Unique3D | Multi-view Consistency9.26 | 8 | 4d ago | |
| Google Scanned Objects (GSO) | TriMM | CLIP Similarity52.5 | 7 | 4d ago | |
| 100 diverse scene images GPT-4o & GPT-Image-1 | EvoScene | Human Pref Win Rate - Geometry Quality85 | 6 | 4d ago | |
| Image-to-3D (Evaluation Set) | Ours | CLIP Score (N)0.758 | 6 | 4d ago | |
| Synthetic 3D Objects (test) | VD | Ewarp0.0021 | 6 | 4d ago | |
| RF15 | DTC123 | CLIP-Similarity0.87 | 4 | 4d ago | |
| Curated dataset from LLaVA-OneVision, Trellis-500K, and Objaverse++ (eval) | CG-MLLM | p-FID12.55 | 3 | 4d ago | |
| GSO | DTC123 | Chamfer Distance0.028 | 3 | 4d ago | |
| RealFusion15 | DreamGaussian | CLIP Similarity0.7 | 3 | 4d ago | |
| Hunyuan3D 3.0 (test) | E-DiT | ULIP-I0.1473 | 2 | 4d ago | |
| Cap3D 2K (test) | Direct image-to-3D | FID32.6 | 1 | 4d ago |
