Depth Estimation on KITTI
0.209RMSEVA-Depth
Evaluation Results
| Method | Links | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| VA-DepthBackbone=Swin-L2026.03 | 0.209 | — | — | — | — | — | — | — | — | — | 97.7 | 99.7 | — | 6.2 | — | — | |
| Depth EstimateBackbone=Swin-L2026.03 | 0.216 | — | — | — | — | — | — | — | — | — | 97.5 | 99.7 | — | 13.3 | — | — | |
| LRHPerceptionBackbone=Swin-m2026.03 | 0.229 | — | — | — | — | — | — | — | — | — | 96.6 | 99.6 | — | 42 | — | — | |
| AdaBinsBackbone=EffNet&Vit2026.03 | 0.236 | — | — | — | — | — | — | — | — | — | 96.4 | 99.5 | — | 1.7 | — | — | |
| ASTransBackbone=Vit-B2026.03 | 0.269 | — | — | — | — | — | — | — | — | — | 96.3 | 99.5 | — | — | — | — | |
| DORNBackbone=ResNet2026.03 | 0.273 | — | — | — | — | — | — | — | — | — | 93.2 | 98.4 | — | — | — | — | |
| BTSBackbone=DenseNet2026.03 | 0.28 | — | — | — | — | — | — | — | — | — | 95.5 | 99.3 | — | 20.1 | — | — | |
| UniDepthV2Evaluation Protocol=zero-shot2026.02 | 1.71 | 0.037 | — | 0.061 | 98.9 | 99.8 | 99.9 | — | — | — | — | — | — | — | — | — | |
| RAD-LargeEvaluation Protocol=zero-shot, Model Scale=Large2026.02 | 1.801 | 0.043 | — | 0.065 | 98.9 | 99.8 | 100 | — | — | — | — | — | — | — | — | — | |
| DepthAnyV2Evaluation Protocol=fine-tuning2026.02 | 1.861 | 0.045 | — | 0.067 | 98.3 | 99.8 | 100 | — | — | — | — | — | — | — | — | — | |
| RAD-BaseEvaluation Protocol=zero-shot, Model Scale=Base2026.02 | 1.913 | 0.046 | — | 0.07 | 98.1 | 99.8 | 100 | — | — | — | — | — | — | — | — | — | |
| Metric3Dv2Evaluation Protocol=zero-shot2026.02 | 1.99 | 0.044 | — | 0.064 | 98.5 | 99.8 | 100 | — | — | — | — | — | — | — | — | — | |
| IEBinsEvaluation Protocol=fine-tuning2026.02 | 2.011 | 0.05 | — | 0.075 | 97.1 | 99.6 | 99.9 | — | — | — | — | — | — | — | — | — | |
| iDiscEvaluation Protocol=fine-tuning2026.02 | 2.07 | 0.05 | — | 0.077 | 97.5 | 99.7 | 99.9 | — | — | — | — | — | — | — | — | — | |
| Li et al.Note=State-of-the-art reference2023.04 | 2.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-g/14, Evaluation Protocol=DPT2023.04 | 2.11 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| NeWCRFsEvaluation protocol=Supervised2023.07 | 2.129 | 0.052 | — | 0.079 | 97.4 | 99.7 | 99.9 | — | — | — | — | — | — | — | — | — | |
| NewCRFEvaluation Protocol=fine-tuning2026.02 | 2.129 | 0.052 | — | 0.079 | 97.4 | 99.7 | 99.9 | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-L/14, Evaluation Protocol=DPT2023.04 | 2.14 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-B/14, Evaluation Protocol=DPT2023.04 | 2.23 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ZoeDepthEvaluation Protocol=fine-tuning2026.02 | 2.281 | 0.054 | — | 0.082 | 97.1 | 99.6 | 99.9 | — | — | — | — | — | — | — | — | — | |
| ProMotion2024.06 | 2.298 | 0.054 | 0.182 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RAD-SmallEvaluation Protocol=zero-shot, Model Scale=Small2026.02 | 2.311 | 0.051 | — | 0.077 | 97.7 | 99.7 | 99.9 | — | — | — | — | — | — | — | — | — | |
| DPT-Hybrid2024.06 | 2.315 | 0.059 | 0.19 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-S/14, Evaluation Protocol=DPT2023.04 | 2.34 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3Param.=7B, Patch Size=16, Evaluation Protocol=Linear probe, Input Resolution=Default2026.03 | 2.346 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3Number of Parameters=7B, Patch Size=16, Evaluation Protocol=Linear probing on frozen features, Evaluation Resolution=default2026.07 | 2.346 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-g/14, Evaluation Protocol=lin. 42023.04 | 2.35 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AdabinsEvaluation protocol=Supervised2023.07 | 2.36 | 0.058 | — | 0.088 | 96.4 | 99.5 | 99.9 | — | — | — | — | — | — | — | — | — | |
| AdaBins2024.06 | 2.36 | 0.058 | 0.198 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AdaBinsEvaluation Protocol=fine-tuning2026.02 | 2.36 | 0.058 | — | 0.088 | 96.4 | 99.5 | 99.9 | — | — | — | — | — | — | — | — | — | |
| BTSEvaluation Protocol=fine-tuning2026.02 | 2.43 | 0.056 | — | 0.089 | 96.2 | 99.4 | 99.8 | — | — | — | — | — | — | — | — | — | |
| V-JEPA 2.1 ViT-GParam.=2B, Patch Size=16, Evaluation Protocol=Linear probe, Input Resolution=Default2026.03 | 2.461 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| V-JEPA 2.1 ViT-GNumber of Parameters=2B, Patch Size=16, Evaluation Protocol=Linear probing on frozen features, Evaluation Resolution=default2026.07 | 2.461 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-LResolution=7682026.05 | 2.4821 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-L/14, Evaluation Protocol=lin. 42023.04 | 2.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| P3Depth2024.06 | 2.519 | 0.06 | 0.206 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-LResolution=5122026.05 | 2.5346 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| iBOTArch=ViT-L/16, Evaluation Protocol=DPT2023.04 | 2.55 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LingBot-Vision ViT-gNumber of Parameters=1B, Patch Size=16, Evaluation Protocol=Linear probing on frozen features, Evaluation Resolution=default2026.07 | 2.552 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OpenCLIPArch=ViT-G/14, Evaluation Protocol=DPT2023.04 | 2.56 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-LResolution=7682026.05 | 2.5634 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LingBot-VisionSize=L2026.07 | 2.574 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MAEArch=ViT-H/14, Evaluation Protocol=DPT2023.04 | 2.59 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-B/14, Evaluation Protocol=lin. 42023.04 | 2.59 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-LResolution=5122026.05 | 2.5904 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| V-JEPA 2.1 ViT-gParam.=1B, Patch Size=16, Evaluation Protocol=Linear probe, Input Resolution=Default2026.03 | 2.601 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| V-JEPA 2.1 ViT-gNumber of Parameters=1B, Patch Size=16, Evaluation Protocol=Linear probing on frozen features, Evaluation Resolution=default2026.07 | 2.601 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AM-RADIOv2.5-B/16Protocol=Linear probing, Backbone=B/162026.05 | 2.6051 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| V-JEPA2.1Size=L2026.07 | 2.617 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-g/14, Evaluation Protocol=lin. 12023.04 | 2.62 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Param.=1B, Patch Size=14, Evaluation Protocol=Linear probe, Input Resolution=Default2026.03 | 2.624 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Number of Parameters=1B, Patch Size=14, Evaluation Protocol=Linear probing on frozen features, Evaluation Resolution=default2026.07 | 2.624 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3 ViT-H+Param.=0.8B, Patch Size=16, Evaluation Protocol=Linear probe, Input Resolution=Default2026.03 | 2.635 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3 ViT-H+Number of Parameters=0.8B, Patch Size=16, Evaluation Protocol=Linear probing on frozen features, Evaluation Resolution=default2026.07 | 2.635 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3Size=L2026.07 | 2.643 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-BResolution=7682026.05 | 2.684 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-BResolution=5122026.05 | 2.6915 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-B/16Protocol=Linear probing, Backbone=B/162026.05 | 2.6915 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2-B/14Protocol=Linear probing, Backbone=B/142026.05 | 2.7002 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-BResolution=5122026.05 | 2.7252 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-B/16Protocol=Linear probing, Backbone=B/16, Core budget (C)=642026.05 | 2.7252 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DORN2019.07 | 2.727 | 0.072 | — | 0.12 | 93.2 | 98.4 | 99.4 | — | — | — | — | — | — | — | — | — | |
| DINOArch=ViT-B/8, Evaluation Protocol=DPT2023.04 | 2.74 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-BResolution=7682026.05 | 2.7468 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TrDepthEvaluation protocol=Supervised2023.07 | 2.755 | 0.064 | — | 0.098 | 95.6 | 99.4 | 99.9 | — | — | — | — | — | — | — | — | — | |
| TransDepth2024.06 | 2.755 | 0.064 | 0.252 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CSTM_labelEvaluation protocol=Zero-shot2023.07 | 2.77 | 0.058 | — | 0.092 | 96.4 | 99.3 | 99.8 | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-L/14, Evaluation Protocol=lin. 12023.04 | 2.78 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-LResolution=3842026.05 | 2.7836 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LingBot-VisionSize=B2026.07 | 2.793 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-LResolution=3842026.05 | 2.8019 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-S+Resolution=7682026.05 | 2.8125 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-S+Resolution=5122026.05 | 2.8127 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3Size=B2026.07 | 2.826 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2-reg-B/14Protocol=Linear probing, Backbone=B/142026.05 | 2.8307 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CSTM_imageEvaluation protocol=Zero-shot2023.07 | 2.843 | 0.06 | — | 0.087 | 96.7 | 99.5 | 99.9 | — | — | — | — | — | — | — | — | — | |
| DINOv3Size=S2026.07 | 2.851 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-S/14, Evaluation Protocol=lin. 42023.04 | 2.86 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-LResolution=2562026.05 | 2.8792 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| V-JEPA2.1Size=B2026.07 | 2.883 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-SResolution=7682026.05 | 2.8943 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-BResolution=3842026.05 | 2.8962 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-BResolution=3842026.05 | 2.8976 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-B/14, Evaluation Protocol=lin. 12023.04 | 2.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-SResolution=5122026.05 | 2.9083 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AM-RADIOv2.5Param.=1B, Patch Size=14, Evaluation Protocol=Linear probe, Input Resolution=Default2026.03 | 2.918 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| AM-RADIOv2.5Number of Parameters=1B, Patch Size=14, Evaluation Protocol=Linear probing on frozen features, Evaluation Resolution=default2026.07 | 2.918 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-LResolution=2562026.05 | 2.9207 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-BResolution=2562026.05 | 2.9636 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-BResolution=2562026.05 | 2.9765 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-S+Resolution=5122026.05 | 2.9878 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv3-S+Resolution=3842026.05 | 2.9987 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-SResolution=5122026.05 | 3.013 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-S+Resolution=7682026.05 | 3.0176 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VECA-SResolution=7682026.05 | 3.0289 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| iBOTArch=ViT-L/16, Evaluation Protocol=lin. 42023.04 | 3.07 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PEspatialParam.=2B, Patch Size=14, Evaluation Protocol=Linear probe, Input Resolution=Default2026.03 | 3.082 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PEspatialNumber of Parameters=2B, Patch Size=14, Evaluation Protocol=Linear probing on frozen features, Evaluation Resolution=default2026.07 | 3.082 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOv2Arch=ViT-S/14, Evaluation Protocol=lin. 12023.04 | 3.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |