Depth Completion on 7-Scenes (AbsRel)
2.6AbsRel (SfM)CAPAVPT
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CAPAVPTBackbone=VGGT, Adaptation strategy=VPT2026.02 | 2.6 | 1.1 | 1.1 | |
| CAPALoRABackbone=VGGT, Adaptation strategy=LoRA2026.02 | 2.8 | 1 | 0.9 | |
| TestPromptDCEstimator type=Depth completion2026.02 | 3.1 | 4 | 2.6 | |
| OMNI-DC-v1.1Estimator type=Depth completion2026.02 | 3.3 | 2.2 | 4.6 | |
| PriorDA-v1.1Estimator type=Depth completion2026.02 | 3.7 | 3.1 | 9.1 | |
| MoGe-2Estimator type=Single-image depth estimation2026.02 | 3.9 | 3.3 | 3.3 | |
| Marigold-DCEstimator type=Depth completion2026.02 | 4.1 | 4 | 1.6 | |
| UniDepthV2Estimator type=Single-image depth estimation2026.02 | 4.2 | 3.7 | 3.7 | |
| VGGTEstimator type=Multi-view reconstruction2026.02 | 4.7 | 4.2 | 4.2 | |
| DepthAnythingV2Estimator type=Single-image depth estimation2026.02 | 5.4 | 4.6 | 4.6 | |
| VideoDAEstimator type=Video-based depth estimation2026.02 | 6.1 | 5.1 | 5.1 | |
| PromptDAEstimator type=Depth completion2026.02 | 8.2 | 6.9 | 4 |