Table 2 Quantitative evaluation of the KITTI dataset with various monocular depth estimation methods.

Method	RMSE	RMSLE	Abs Rel	Sq Rel	δ < 1.25	δ < 1.25²	δ < 1.25³
Method	Low is better				High is better
Eigen et al.⁵	7.156	0.270	0.190	1.515	0.692	0.899	0.967
Kuznietsov et al.¹⁰	3.610	0.138	0.113	0.478	0.906	0.980	0.995
Godard et al.⁴⁴	4.630	0.193	0.106	0.806	0.876	0.958	0.980
Gan et al.⁴⁵	3.933	0.173	0.098	–	0.890	0.964	0.985
Xu et al.⁴⁶	3.842	0.185	0.092	–	0.895	0.974	0.990
Bae et al.³³	3.457	0.113	0.071	0.436	0.939	0.987	0.996
Gonzalez Bello et al.³⁶	2.988	0.107	0.070	0.285	0.946	0.991	0.998
Wang et al.⁴⁷	2.896	0.097	0.058	0.286	0.959	0.992	0.998
Fu et al.²⁹	2.727	0.120	0.072	0.307	0.932	0.984	0.995
Lee et al. ⁴⁸	2.756	0.096	0.059	0.241	0.956	0.993	0.998
Song et al.³⁰	2.446	0.091	0.059	0.212	0.962	0.993	0.999
Ours	2.247	0.085	0.055	0.200	0.964	0.994	0.999

Quick links

Search