Table 1 Depth estimation result comparison with state-of-the-art models on the NYU dataset.

From: Monocular depth estimation via a detail semantic collaborative network for indoor scenes

Method

Accuracy (Higher is better)

Error (Lower is better)

δ < 1.25

δ < 1.252

δ < 1.253

REL

RMSE

Log10

Eigen et al.28

0.611

0.887

0.971

0.215

0.907

-

Wang et al.23

0.605

0.890

0.970

0.220

0.745

0.094

Laina et al.31

0.811

0.953

0.988

0.127

0.573

0.055

Mousavian et al.24

0.568

0.856

0.956

-

0.816

0.061

Xu et al.25

0.811

0.954

0.987

0.121

0.586

0.052

Fu et al.29

0.828

0.965

0.992

0.115

0.509

0.051

Hu et al.26

0.866

0.975

0.993

0.115

0.530

0.050

Chen et al.39

0.878

0.977

0.994

0.111

0.514

0.048

Yu et al.27

0.772

0.942

0.984

0.159

0.599

0.068

Huynh et al.17

0.882

0.980

0.996

0.108

0.412

-

Bhat et al.18

0.903

0.984

0.997

0.103

0.364

0.044

Vaishakh et al.68

0.902

0.984

0.997

0.101

0.353

0.042

Kim et al.19

0.915

0.988

0.997

0.098

0.344

0.042

Vitor et al.69

0.895

0.965

-

0.104

0.389

-

Ours

0.916

0.988

0.997

0.097

0.342

0.041

Back to article page