NYU Depth
Benchmarks
TaskDatasetSOTA resultTrendResults
NYU Depth v2
0.119RMSE
226
NYU-depth-v2 official (test)
0.085RMSE
200
NYU Depth V2 (test)
61.2mIoU
183
NYU-depth v2 (test)
99.5Accuracy (δ < 1.25)
113
NYU-Depth v2 (official)
0.056Abs Rel
97
NYU Depth Eigen v2 (test)
0.044A.Rel
57
NYU Depth V2
68.1mIoU
56
NYU-Depth v2
0.089RMSE
43
NYU-Depth V2 (test)
79.4Precision
30
NYU-depth v2
58.8mIoU
28
NYU-Depth v2 (val)
3.9A.Rel
27
NYU-Depth V2 (test)
0.041AbsRel
24
NYU Depth V2
0.641Recall
18
NYU Depth V2
73.12Accuracy (Clean)
17
NYU Depth v2 (test)
55.4mAP
17
NYU-Depth Setting C: T* -> T V2 (test)
0.09RMSE
11
NYU Depth (test)
71.62AUC @10°
11
NYU Depth v2 (Underrepresented classes)
97.5Delta 1
10
NYU-Depth V2
33.5RGB-to-D Correspondence Score
10
NYU Depth v2 (test)
70.4Mean Class Accuracy
9
NYU-Depth Setting A: R -> T V2 (test)
0.12RMSE
9
NYU-Depth D-HAZY (full)
15.54PSNR
9
NYU-Depth Setting B: R* -> T V2 (test)
0.242RMSE
8
NYU-Depth Setting C: T* ⇒ T V2 (test)
47.39CD
7
NYU-Depth Setting B: R* ⇒ T V2 (test)
249.05CD
7