Monocular Depth Estimation on ETH-3D (test)
0.048A.RelLAGRNet
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| LAGRNeteval_mode=Zero-shot evaluation2026.04 | 0.048 | — | 98.9 | — | |
| DistillAnyDeptheval_mode=Zero-shot evaluation2026.04 | 0.054 | — | 98.1 | — | |
| GRINzero-shot=true, test-time scale alignment=true, intrinsics required=false2024.09 | 0.061 | — | — | — | |
| CSTM_imageBackbone=ConvNeXt-large, #Params=198M, Zero-shot transfer=true, Manually aligned scale and shift=true2023.07 | 0.064 | — | 96.5 | — | |
| Marigoldzero-shot=true, test-time scale alignment=true, intrinsics required=false2024.09 | 0.065 | — | — | — | |
| Marigoldeval_mode=Zero-shot evaluation2026.04 | 0.065 | — | 96 | — | |
| CSTM_labelBackbone=ConvNeXt-large, Zero-shot transfer=true, Manually aligned scale and shift=true2023.07 | 0.066 | — | 96.4 | — | |
| DPT-largeBackbone=ViT-Large, Zero-shot transfer=true, Manually aligned scale and shift=true2023.07 | 0.078 | — | 94.6 | — | |
| DPTeval_mode=Zero-shot evaluation2026.04 | 0.078 | — | 94.6 | — | |
| GenPercepteval_mode=Zero-shot evaluation2026.04 | 0.096 | — | 95.9 | — | |
| HDNBackbone=ViT-Large, #Params=306M, Zero-shot transfer=true, Manually aligned scale and shift=true2023.07 | 0.121 | — | 83.3 | — | |
| HDNeval_mode=Zero-shot evaluation2026.04 | 0.121 | — | 83.3 | — | |
| Depth AnythingEncoder=ViT-B, Zero-shot=true2024.01 | 0.126 | — | 88.4 | — | |
| DepthAnythingzero-shot=true, test-time scale alignment=true, intrinsics required=false2024.09 | 0.126 | — | — | — | |
| Depth Anything V1Encoder=ViT-B, Zero-shot=true2024.06 | 0.126 | — | 88.4 | — | |
| Depth AnythingEncoder=ViT-S, Zero-shot=true2024.01 | 0.127 | — | 88.5 | — | |
| Depth AnythingEncoder=ViT-L, Zero-shot=true2024.01 | 0.127 | — | 88.2 | — | |
| Depth Anything V1Encoder=ViT-S, Zero-shot=true2024.06 | 0.127 | — | 88.5 | — | |
| Depth Anything V1Encoder=ViT-L, Zero-shot=true2024.06 | 0.127 | — | 88.2 | — | |
| DepthAnything v1eval_mode=Zero-shot evaluation2026.04 | 0.127 | — | 88.2 | — | |
| Depth Anything V2Encoder=ViT-L, Zero-shot=true2024.06 | 0.131 | — | 86.5 | — | |
| DepthAnything v2eval_mode=Zero-shot evaluation2026.04 | 0.131 | — | 86.5 | — | |
| Depth Anything V2Encoder=ViT-G, Zero-shot=true2024.06 | 0.132 | — | 86.2 | — | |
| Depth Anything V2Encoder=ViT-B, Zero-shot=true2024.06 | 0.137 | — | 85.8 | — | |
| MiDaS v3.1Encoder=ViT-L, Zero-shot=true2024.01 | 0.139 | — | 86.7 | — | |
| MiDaS V3.1Encoder=ViT-L, Zero-shot=true2024.06 | 0.139 | — | 86.7 | — | |
| Depth Anything V2Encoder=ViT-S, Zero-shot=true2024.06 | 0.142 | — | 85.1 | — | |
| OmnidataBackbone=ViT-base, Zero-shot transfer=true, Manually aligned scale and shift=true2023.07 | 0.166 | — | 77.8 | — | |
| Omnidataeval_mode=Zero-shot evaluation2026.04 | 0.166 | — | 77.8 | — | |
| LeresBackbone=ResNeXt101, Zero-shot transfer=true, Manually aligned scale and shift=true2023.07 | 0.171 | — | 77.7 | — | |
| MiDaSBackbone=ResNeXt101, #Params=88M, Zero-shot transfer=true, Manually aligned scale and shift=true2023.07 | 0.184 | — | 75.2 | — | |
| MiDaSeval_mode=Zero-shot evaluation2026.04 | 0.184 | — | 75.2 | — | |
| DiverseDepthBackbone=ResNeXt50, #Params=25M, Zero-shot transfer=true, Manually aligned scale and shift=true2023.07 | 0.228 | — | 69.4 | — | |
| DiverseDeptheval_mode=Zero-shot evaluation2026.04 | 0.228 | — | 69.4 | — | |
| VNLTraining Dataset=NYU Depth-v2, Zero-shot=true2022.04 | 0.331 | 1.649 | 46.2 | — | |
| P3DepthTraining Dataset=NYU Depth-v2, Zero-shot=true2022.04 | 0.343 | 1.637 | 46.8 | — | |
| BTSTraining Dataset=NYU Depth-v2, Zero-shot=true2022.04 | 0.366 | 1.84 | 39.8 | — | |
| BetterDepthZero-shot=true2024.07 | 4.7 | — | 98.1 | — | |
| Depth AnythingZero-shot=true2024.07 | 6.2 | — | 98 | — | |
| Depth Anything V2Zero-shot=true2024.07 | 6.2 | — | 98.2 | — | |
| Marigoldzero-shot=true, Num. Train Samples=74K, Architecture Type=Multi-step diffusion2024.09 | 6.5 | — | 96 | 2.3 | |
| MarigoldZero-shot=true2024.07 | 6.5 | — | 96 | — | |
| PrimeDepthzero-shot=true, Num. Train Samples=74K, Architecture Type=Single-step diffusion2024.09 | 6.8 | — | 96.7 | 2 | |
| DPTzero-shot=true, Num. Train Samples=1.4M, Architecture Type=Transformer2024.09 | 7.8 | — | 94.6 | 4.7 | |
| HDNzero-shot=true, Num. Train Samples=300K, Architecture Type=Transformer2024.09 | 12.1 | — | 83.3 | 4.5 | |
| Depth Anythingzero-shot=true, Num. Train Samples=63.5M, Architecture Type=Transformer2024.09 | 12.7 | — | 88.2 | 2.2 | |
| Omnidatazero-shot=true, Num. Train Samples=12.2M, Architecture Type=Transformer2024.09 | 16.6 | — | 77.8 | 5.7 | |
| LeReSzero-shot=true, Num. Train Samples=354K, Architecture Type=ResNeXt1012024.09 | 17.1 | — | 77.7 | 6.5 | |
| MiDaSzero-shot=true, Num. Train Samples=2M, Architecture Type=ResNeXt1012024.09 | 18.4 | — | 75.2 | 8 |