Table 3 Quantitative results for zero-shot transfer to three unseen indoor datasets and one outdoor dataset.
From: A simple monocular depth estimation network for balancing complexity and accuracy
Method | SUN RGB-D | iBims-1 Benchmark | ||||
|---|---|---|---|---|---|---|
\(\delta _1\uparrow\) | AbsRel\(\downarrow\) | RMSE\(\downarrow\) | \(\delta _1\uparrow\) | AbsRel\(\downarrow\) | RMSE\(\downarrow\) | |
BTS58 | 0.740 | 0.172 | 0.515 | 0.538 | 0.231 | 0.919 |
AdaBins12 | 0.771 | 0.159 | 0.476 | 0.555 | 0.212 | 0.901 |
LocalBins66 | 0.797 | 0.151 | 0.424 | 0.548 | 0.206 | 0.861 |
NeWCRFs19 | 0.777 | 0.156 | 0.470 | 0.558 | 0.211 | 0.880 |
IEBins18 | 0.768 | 0.157 | 0.476 | 0.601 | 0.204 | 0.849 |
Ours | 0.781 | 0.151 | 0.423 | 0.578 | 0.197 | 0.831 |
Method | DIODE Indoor | DIODE outdoor | ||||
|---|---|---|---|---|---|---|
\(\delta _1\uparrow\) | AbsRel\(\downarrow\) | RMSE\(\downarrow\) | \(\delta _1\uparrow\) | AbsRel\(\downarrow\) | RMSE\(\downarrow\) | |
BTS58 | 0.210 | 0.418 | 1.905 | 0.171 | 0.837 | 10.480 |
AdaBins12 | 0.174 | 0.443 | 1.963 | 0.161 | 0.863 | 10.350 |
LocalBins66 | 0.229 | 0.412 | 1.853 | 0.170 | 0.821 | 10.270 |
NeWCRFs19 | 0.187 | 0.404 | 1.867 | 0.168 | 0.759 | 9.228 |
IEBins18 | 0.231 | 0.387 | 1.787 | 0.128 | 0.685 | 9.360 |
Ours | 0.270 | 0.366 | 1.717 | 0.188 | 0.711 | 8.691 |
