Monocular Depth Estimation on VOID (test)
78.2Accuracy Threshold (< 1.25)DPT
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| DPTScaling=Median, Training Dataset=VOID2024.10 | 78.2 | 96.2 | 99 | 0.15 | 0.064 | 0.34 | |
| RSABackbone=DPT, Training Dataset=NYUv2, KITTI, zero-shot=true2024.10 | 60.1 | 88.6 | 97 | 0.254 | 0.096 | 0.444 | |
| RSABackbone=DPT, Training Dataset=NYUv2, KITTI, VOID2024.10 | 59.8 | 87.7 | 95.6 | 0.248 | 0.1 | 0.475 | |
| RSABackbone=MiDas, Training Dataset=NYUv2, KITTI, VOID2024.10 | 53.5 | 82.9 | 94.5 | 0.28 | 0.112 | 0.528 | |
| DPTScaling=Image, Training Dataset=NYUv2, KITTI, VOID2024.10 | 53.4 | 82.7 | 94.1 | 0.266 | 0.108 | 0.545 | |
| DPTScaling=Image, Training Dataset=NYUv2, KITTI, zero-shot=true2024.10 | 51.6 | 81.2 | 93.6 | 0.289 | 0.112 | 0.634 | |
| MiDasScaling=Median, Training Dataset=VOID2024.10 | 50 | 78.1 | 89.9 | 0.347 | 0.13 | 0.829 | |
| DPTScaling=Global, Training Dataset=NYUv2, zero-shot=true2024.10 | 45.6 | 74.3 | 91.2 | 0.312 | 0.136 | 0.896 | |
| RSABackbone=MiDas, Training Dataset=NYUv2, KITTI, zero-shot=true2024.10 | 39.2 | 69.6 | 89.2 | 0.448 | 0.148 | 0.66 | |
| MiDasScaling=Image, Training Dataset=NYUv2, KITTI, VOID2024.10 | 38.9 | 74.3 | 91.1 | 0.392 | 0.139 | 0.652 | |
| RSABackbone=DepthAnything, Training Dataset=NYUv2, KITTI, VOID2024.10 | 37.4 | 67.3 | 83.7 | 0.477 | 0.168 | 0.792 | |
| DepthAnythingScaling=Image, Training Dataset=NYUv2, KITTI, VOID2024.10 | 32.3 | 61.2 | 76.8 | 0.589 | 0.196 | 0.956 | |
| MiDasScaling=Image, Training Dataset=NYUv2, KITTI, zero-shot=true2024.10 | 30.4 | 62.6 | 81.2 | 0.487 | 0.159 | 0.913 | |
| MiDasScaling=Global, Training Dataset=NYUv2, zero-shot=true2024.10 | 26.8 | 59.7 | 73.5 | 0.512 | 0.193 | 1.346 | |
| DepthAnythingScaling=Median, Training Dataset=VOID2024.10 | 24.9 | 46.5 | 64.3 | 0.682 | 0.254 | 1.251 | |
| RSABackbone=DepthAnything, Training Dataset=NYUv2, KITTI, zero-shot=true2024.10 | 10.4 | 26.2 | 45 | 1.287 | 0.323 | 1.716 | |
| DepthAnythingScaling=Image, Training Dataset=NYUv2, KITTI, zero-shot=true2024.10 | 9.3 | 21.5 | 41.2 | 1.497 | 0.345 | 1.963 | |
| DepthAnythingScaling=Global, Training Dataset=NYUv2, zero-shot=true2024.10 | 8.4 | 19.4 | 37.6 | 1.674 | 0.389 | 2.046 |