Monocular Depth Estimation on KITTI-360 (test)
2.04RMSEUniDepth + Calibration Tokens
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| UniDepth + Calibration TokensExperiment=+ Calibration Tokens, Model=UniDepth, Train Dataset=Mix 200K, Zero-shot=true2025.08 | 2.04 | 66.4 | |
| DepthAnything + Calibration TokensExperiment=+ Calibration Tokens, Model=DepthAnything, Train Dataset=Mix 200K, Zero-shot=true2025.08 | 2.043 | 81 | |
| DepthAnyCameraExperiment=Comparisons, Model=DepthAnyCamera, Train Dataset=Outdoor 130K, Zero-shot=true2025.08 | 2.067 | 85.2 | |
| UniDepthExperiment=Baseline, Model=UniDepth, Train Dataset=Mix 16M, Zero-shot=true2025.08 | 2.085 | 66.3 | |
| DepthAnythingExperiment=Baseline, Model=DepthAnything, Train Dataset=Mix 63.5M, Zero-shot=true2025.08 | 2.214 | 83.9 | |
| MiDaS + Calibration TokensExperiment=+ Calibration Tokens, Model=MiDaS, Train Dataset=Mix 200K, Zero-shot=true2025.08 | 2.348 | 65.8 | |
| FoVA-DepthExperiment=Comparisons, Model=FoVA-Depth, Train Dataset=Outdoor 80K, Zero-shot=true2025.08 | 3.096 | 63.2 | |
| MiDaSExperiment=Baseline, Model=MiDaS, Train Dataset=Mix 1.4M, Zero-shot=true2025.08 | 3.312 | 58.6 | |
| DepthAnyCameraExperiment=Comparisons, Model=DepthAnyCamera, Train Dataset=Mix 200K, Zero-shot=true2025.08 | 5.675 | 34.8 |