Depth Estimation on ScanNet
0.023AbsRelMetric3D v2
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Metric3D v2Training Samples=16M, NFEs=-, Zero-shot=true2025.12 | 0.023 | — | — | 98.9 | 1.2 | — | — | — | — | — | |
| VGGTScale factor alignment=optimized shared scale factor across the entire video, trained_on_dataset=ScanNet2026.06 | 0.029 | — | — | 98.9 | — | — | — | — | — | 0.048 | |
| DepthAnything v2Training Samples=62.6M, NFEs=-, Zero-shot=true2025.12 | 0.042 | — | — | 97.9 | 2.8 | — | — | — | — | — | |
| Depth Prosetting=zero-shot, seeds=metric depth estimator2026.05 | 0.042 | — | — | 98 | — | — | — | — | — | — | |
| DepthAnythingTraining Samples=63.5K, NFEs=-, Zero-shot=true2025.12 | 0.043 | — | — | 98.1 | 2.2 | — | — | — | — | — | |
| DepthAnythingV2setting=zero-shot, seeds=metric depth estimator2026.05 | 0.043 | — | — | 98.1 | — | — | — | — | — | — | |
| MTDsetting=zero-shot, seeds=metric depth estimator2026.05 | 0.044 | — | — | 97.8 | — | — | — | — | — | — | |
| DepthMastersetting=zero-shot, seeds=metric depth estimator2026.05 | 0.055 | — | — | 96.7 | — | — | — | — | — | — | |
| GenperceptTraining Samples=90K, NFEs=1 x 1, Zero-shot=true2025.12 | 0.056 | — | — | 96.5 | 4 | — | — | — | — | — | |
| E2E-FTTraining Samples=74K, NFEs=1 x 1, Zero-shot=true2025.12 | 0.058 | — | — | 96.5 | 3.1 | — | — | — | — | — | |
| Lotus-GTraining Samples=59K, NFEs=1 x 1, Zero-shot=true2025.12 | 0.06 | — | — | 96 | 3.3 | — | — | — | — | — | |
| Lotussetting=zero-shot, seeds=metric depth estimator2026.05 | 0.06 | — | — | 96.3 | — | — | — | — | — | — | |
| GeoWizardsetting=zero-shot, seeds=metric depth estimator2026.05 | 0.061 | — | — | 95.3 | — | — | — | — | — | — | |
| Unified-IOTraining Samples=48K, NFEs=-, Zero-shot=true2025.12 | 0.063 | — | — | 96.5 | 1.9 | — | — | — | — | — | |
| D3-PredictorTraining Samples=30K, NFEs=1 x 1, Zero-shot=true2025.12 | 0.063 | — | — | 96.3 | 2.1 | — | — | — | — | — | |
| MarigoldTraining Samples=74K, NFEs=10 x 50, Zero-shot=true2025.12 | 0.064 | — | — | 95.1 | 4.4 | — | — | — | — | — | |
| Marigoldsetting=zero-shot, seeds=metric depth estimator2026.05 | 0.064 | — | — | 95.1 | — | — | — | — | — | — | |
| 4M-XLTraining Samples=759M, NFEs=-, Zero-shot=true2025.12 | 0.065 | — | — | 95.5 | 1.7 | — | — | — | — | — | |
| GeoWizardTraining Samples=280K, NFEs=10 x 50, Zero-shot=true2025.12 | 0.066 | — | — | 95.3 | 5.6 | — | — | — | — | — | |
| OmnidataTraining Samples=12.2M, NFEs=-, Zero-shot=true2025.12 | 0.075 | — | — | 93.6 | 5.1 | — | — | — | — | — | |
| DeceptionTraining Samples=500K, NFEs=1 x 28, Zero-shot=true2025.12 | 0.075 | — | — | 93.8 | 3.6 | — | — | — | — | — | |
| Omnidatasetting=zero-shot, seeds=metric depth estimator2026.05 | 0.075 | — | — | 93.6 | — | — | — | — | — | — | |
| HDNTraining Samples=300K, NFEs=-, Zero-shot=true2025.12 | 0.08 | — | — | 93.9 | 3.9 | — | — | — | — | — | |
| HDNsetting=zero-shot, seeds=metric depth estimator2026.05 | 0.08 | — | — | 93.9 | — | — | — | — | — | — | |
| SCOPEScale factor alignment=optimized shared scale factor across the entire video2026.06 | 0.081 | — | — | 94.1 | — | — | — | — | — | 0.049 | |
| DPT-largeTraining Samples=1.4M, NFEs=-, Zero-shot=true2025.12 | 0.082 | — | — | 93.4 | 4.2 | — | — | — | — | — | |
| DPTsetting=zero-shot, seeds=metric depth estimator2026.05 | 0.084 | — | — | 93.2 | — | — | — | — | — | — | |
| DepthProScale factor alignment=optimized shared scale factor across the entire video2026.06 | 0.089 | — | — | 92.9 | — | — | — | — | — | 0.065 | |
| AFNettrained_on=DDAD, evaluation_mode=zero-shot2024.03 | 0.091 | 0.039 | 0.253 | — | — | — | — | — | — | — | |
| LeReSsetting=zero-shot, seeds=metric depth estimator2026.05 | 0.091 | — | — | 91.7 | — | — | — | — | — | — | |
| OneDiffusionTraining Samples=500K, NFEs=-, Zero-shot=true2025.12 | 0.094 | — | — | 90.6 | 2.4 | — | — | — | — | — | |
| MVS2Dtrained_on=DDAD, evaluation_mode=zero-shot2024.03 | 0.098 | 0.044 | 0.276 | — | — | — | — | — | — | — | |
| MiDaSsetting=zero-shot, seeds=metric depth estimator2026.05 | 0.099 | — | — | 90.7 | — | — | — | — | — | — | |
| DiverseDepthTraining Samples=320K, NFEs=-, Zero-shot=true2025.12 | 0.109 | — | — | 88.2 | 7 | — | — | — | — | — | |
| DiverseDepthsetting=zero-shot, seeds=metric depth estimator2026.05 | 0.109 | — | — | 88.2 | — | — | — | — | — | — | |
| Splat and DistillBackbone=ViT-Base, Evaluation Protocol=Linear Probing2026.02 | 0.1113 | — | 0.2245 | — | — | — | — | — | — | — | |
| MaGNettrained_on=DDAD, evaluation_mode=zero-shot2024.03 | 0.112 | 0.051 | 0.314 | — | — | — | — | — | — | — | |
| Fit3DBackbone=ViT-Base, Evaluation Protocol=Linear Probing2026.02 | 0.1166 | — | 0.2346 | — | — | — | — | — | — | — | |
| DINOv2Backbone=ViT-Base, Evaluation Protocol=Linear Probing2026.02 | 0.1169 | — | 0.2369 | — | — | — | — | — | — | — | |
| MiDaSTraining Samples=2M, NFEs=-, Zero-shot=true2025.12 | 0.121 | — | — | 84.6 | 6.9 | — | — | — | — | — | |
| IterMVStrained_on=DDAD, evaluation_mode=zero-shot2024.03 | 0.123 | 0.056 | 0.332 | — | — | — | — | — | — | — | |
| Splat and DistillBackbone=ViT-Small, Evaluation Protocol=Linear Probing2026.02 | 0.1266 | — | 0.2555 | — | — | — | — | — | — | — | |
| MEFBackbone=ViT-Base, Evaluation Protocol=Linear Probing2026.02 | 0.1269 | — | 0.2534 | — | — | — | — | — | — | — | |
| MoGeScale factor alignment=optimized shared scale factor across the entire video2026.06 | 0.13 | — | — | 85.2 | — | — | — | — | — | 0.077 | |
| Fit3DBackbone=ViT-Small, Evaluation Protocol=Linear Probing2026.02 | 0.1375 | — | 0.2713 | — | — | — | — | — | — | — | |
| DINOv2Backbone=ViT-Small, Evaluation Protocol=Linear Probing2026.02 | 0.1437 | — | 0.2817 | — | — | — | — | — | — | — | |
| DMPTraining Samples=10K, NFEs=1 x 5, Zero-shot=true2025.12 | 0.146 | — | — | 81.4 | 7.7 | — | — | — | — | — | |
| MEFBackbone=ViT-Small, Evaluation Protocol=Linear Probing2026.02 | 0.1566 | — | 0.3042 | — | — | — | — | — | — | — | |
| Leveling3DInput setting=sparse two-view, Evaluation poses=novel2026.03 | 0.252 | — | 24.68 | 82.6 | — | — | — | 0.0376 | — | — | |
| Difix3D+Input setting=sparse two-view, Evaluation poses=novel2026.03 | 0.286 | — | 30.358 | 78.7 | — | — | — | 0.0452 | — | — | |
| GSFix3DInput setting=sparse two-view, Evaluation poses=novel2026.03 | 0.29 | — | 33.398 | 79.1 | — | — | — | 0.0453 | — | — | |
| GSFixerInput setting=sparse two-view, Evaluation poses=novel2026.03 | 0.296 | — | 31.296 | 78.4 | — | — | — | 0.0462 | — | — | |
| BaselineInput setting=sparse two-view, Evaluation poses=novel2026.03 | 0.372 | — | 58.826 | 70.8 | — | — | — | 0.0472 | — | — | |
| TestPromptDCCondition=< 3m, Noise-free=true2026.02 | 0.9 | — | — | — | — | 2.9 | — | — | — | — | |
| CAPALoRACondition=100, Noise-free=true2026.02 | 0.9 | — | — | — | — | 1.8 | — | — | — | — | |
| CAPAVPTCondition=100, Noise-free=true2026.02 | 0.9 | — | — | — | — | 1.9 | — | — | — | — | |
| CAPALoRACondition=SIFT, Noise-free=true2026.02 | 1 | — | — | — | — | 2 | — | — | — | — | |
| CAPALoRACondition=< 3m, Noise-free=true2026.02 | 1 | — | — | — | — | 2.9 | — | — | — | — | |
| CAPAVPTCondition=< 3m, Noise-free=true2026.02 | 1 | — | — | — | — | 2.7 | — | — | — | — | |
| CAPALoRA + MoGe-2Condition=100, Noise-free=true2026.02 | 1 | — | — | — | — | 2.1 | — | — | — | — | |
| TestPromptDCCondition=100, Noise-free=true2026.02 | 1.1 | — | — | — | — | 2.3 | — | — | — | — | |
| CAPAVPTCondition=SIFT, Noise-free=true2026.02 | 1.1 | — | — | — | — | 2.2 | — | — | — | — | |
| CAPALoRA + UniDepthV2Condition=100, Noise-free=true2026.02 | 1.1 | — | — | — | — | 2.2 | — | — | — | — | |
| CAPAVPT + UniDepthV2Condition=100, Noise-free=true2026.02 | 1.1 | — | — | — | — | 2.1 | — | — | — | — | |
| CAPAVPT + MoGe-2Condition=100, Noise-free=true2026.02 | 1.2 | — | — | — | — | 2.3 | — | — | — | — | |
| CAPALoRA + MoGe-2Condition=SIFT, Noise-free=true2026.02 | 1.3 | — | — | — | — | 2.5 | — | — | — | — | |
| CAPALoRA + MoGe-2Condition=< 3m, Noise-free=true2026.02 | 1.3 | — | — | — | — | 3.4 | — | — | — | — | |
| CAPALoRA + UniDepthV2Condition=< 3m, Noise-free=true2026.02 | 1.3 | — | — | — | — | 3.7 | — | — | — | — | |
| OMNI-DC-v1.1Condition=100, Noise-free=true2026.02 | 1.4 | — | — | — | — | 2.8 | — | — | — | — | |
| CAPALoRA + UniDepthV2Condition=SIFT, Noise-free=true2026.02 | 1.4 | — | — | — | — | 2.5 | — | — | — | — | |
| CAPAVPT + UniDepthV2Condition=SIFT, Noise-free=true2026.02 | 1.4 | — | — | — | — | 2.5 | — | — | — | — | |
| CAPAVPT + UniDepthV2Condition=< 3m, Noise-free=true2026.02 | 1.4 | — | — | — | — | 4.1 | — | — | — | — | |
| TestPromptDCCondition=SIFT, Noise-free=true2026.02 | 1.5 | — | — | — | — | 2.8 | — | — | — | — | |
| CAPAVPT + MoGe-2Condition=SIFT, Noise-free=true2026.02 | 1.5 | — | — | — | — | 2.8 | — | — | — | — | |
| CAPAVPT + MoGe-2Condition=< 3m, Noise-free=true2026.02 | 1.5 | — | — | — | — | 3.9 | — | — | — | — | |
| PriorDA-v1.1Condition=100, Noise-free=true2026.02 | 1.6 | — | — | — | — | 3.3 | — | — | — | — | |
| OMNI-DC-v1.1Condition=SIFT, Noise-free=true2026.02 | 1.8 | — | — | — | — | 3.4 | — | — | — | — | |
| Marigold-DCCondition=100, Noise-free=true2026.02 | 1.8 | — | — | — | — | 3.7 | — | — | — | — | |
| VGGTCondition=100, Noise-free=true2026.02 | 1.9 | — | — | — | — | 4.4 | — | — | — | — | |
| VGGTCondition=< 3m, Noise-free=true2026.02 | 1.9 | — | — | — | — | 4.5 | — | — | — | — | |
| VGGTCondition=SIFT, Noise-free=true2026.02 | 2.2 | — | — | — | — | 4.7 | — | — | — | — | |
| Marigold-DCCondition=< 3m, Noise-free=true2026.02 | 2.2 | — | — | — | — | 7.6 | — | — | — | — | |
| PriorDA-v1.1Condition=SIFT, Noise-free=true2026.02 | 2.5 | — | — | — | — | 4.4 | — | — | — | — | |
| UniDepthV2Condition=100, Noise-free=true2026.02 | 2.7 | — | — | — | — | 5.3 | — | — | — | — | |
| OMNI-DC-v1.1Condition=< 3m, Noise-free=true2026.02 | 2.7 | — | — | — | — | 8.5 | — | — | — | — | |
| UniDepthV2Condition=< 3m, Noise-free=true2026.02 | 2.8 | — | — | — | — | 5.9 | — | — | — | — | |
| Marigold-DCCondition=SIFT, Noise-free=true2026.02 | 3 | — | — | — | — | 5.4 | — | — | — | — | |
| UniDepthV2Condition=SIFT, Noise-free=true2026.02 | 3.2 | — | — | — | — | 6 | — | — | — | — | |
| MoGe-2Condition=100, Noise-free=true2026.02 | 3.3 | — | — | — | — | 6.6 | — | — | — | — | |
| MoGe-2Condition=< 3m, Noise-free=true2026.02 | 3.3 | — | — | — | — | 7.2 | — | — | — | — | |
| MoGe-2Method category=Geometry-specialized2026.07 | 3.4 | — | — | 98.3 | — | — | — | — | — | — | |
| UniSemNumber of View Inputs=82026.03 | 3.73 | — | 12.94 | — | — | — | — | — | — | — | |
| ViewExtrapolatorInput setting=sparse two-view, Evaluation poses=novel2026.03 | 3.794 | — | 120.666 | 32.3 | — | — | — | 0.0646 | — | — | |
| DepthAnythingV2Condition=100, Noise-free=true2026.02 | 3.8 | — | — | — | — | 7.5 | — | — | — | — | |
| SenseNova-VisionMethod category=Generation-based2026.07 | 3.9 | — | — | 98 | — | — | — | — | — | — | |
| MoGe-2Condition=SIFT, Noise-free=true2026.02 | 4 | — | — | — | — | 7.6 | — | — | — | — | |
| UniSemNumber of View Inputs=162026.03 | 4.08 | — | 13.41 | — | — | — | — | — | — | — | |
| VideoDACondition=100, Noise-free=true2026.02 | 4.1 | — | — | — | — | 7.7 | — | — | — | — | |
| VideoDACondition=< 3m, Noise-free=true2026.02 | 4.2 | — | — | — | — | 8.5 | — | — | — | — | |
| DepthAnything V2Method category=Geometry-specialized2026.07 | 4.2 | — | — | 97.8 | — | — | — | — | — | — |