Table 2.

Quantitative results with all variants of Monodepth2 [3] and other self-supervised methods on the KITTI 2015 dataset.

Method	Train Input	Abs Rel	Sq Rel	RMSE	RMSE log	δ < 1.25	δ < 1.252	δ < 1.253
AdaDepth 2018 [10]	D*	0.167	1.257	5.578	0.237	0.771	0.922	0.971
Kuznietsov 2017 [33]	DS	0.113	0.741	4.621	0.189	0.862	0.96	0.986
DVSO 2018 [34]	D*S	0.097	0.734	4.442	0.187	0.888	0.958	0.98
SVSM FT 2018 [15]	DS	0.094	0.626	4.252	0.177	0.891	0.965	0.984
Guo 2018 [32]	DS	0.096	0.641	4.095	0.168	0.892	0.967	0.986
DORN 2018 [35]	D	0.072	0.307	2.727	0.12	0.932	0.984	0.994
Zhou 2017 [28]	M	0.183	1.595	6.709	0.27	0.734	0.902	0.959
Yang 2018 [7]	M	0.182	1.481	6.501	0.267	0.725	0.906	0.963
Mahjourian 2018 [36]	M	0.163	1.24	6.22	0.25	0.762	0.916	0.968
GeoNet 2018 [37]	M	0.149	1.06	5.567	0.226	0.796	0.935	0.975
DDVO 2018 [29]	M	0.151	1.257	5.583	0.228	0.81	0.936	0.974
Ranjan 2019 [38]	M	0.148	1.149	5.464	0.226	0.815	0.935	0.973
EPC++ 2020 [39]	M	0.141	1.029	5.35	0.216	0.816	0.941	0.976
Struct2depth 2019 [40]	M	0.141	1.026	5.291	0.215	0.816	0.945	0.979
Monodepth2 w/o pretraining 2019 [3]	M	0.132	1.044	5.142	0.21	0.845	0.948	0.977
Monodepth2 (640 × 192), 2019 [3]	M	0.115	0.903	4.863	0.193	0.877	0.959	0.981
Monodepth2 (1024 × 320), 2019 [3]	M	0.115	0.882	4.701	0.19	0.879	0.961	0.982
BTS ResNet50, 2019 [11]	M	0.061	0.261	2.834	0.099	0.954	0.992	0.998
Garg 2016 [41]	S	0.152	1.226	5.849	0.246	0.784	0.921	0.967
Monodepth R50 2017 [6]	S	0.133	1.142	5.533	0.23	0.83	0.936	0.97
StrAT 2018 [42]	S	0.128	1.019	5.403	0.227	0.827	0.935	0.971
3Net (VGG), 2018 [43]	S	0.119	1.201	5.888	0.208	0.844	0.941	0.978
SuperDepth & PP, 2019 [44] (1024 × 382)	S	0.112	0.875	4.958	0.207	0.852	0.947	0.977
Monodepth2 w/o pretraining 2019 [3]	S	0.13	1.144	5.485	0.232	0.831	0.932	0.968
Monodepth2 (640 × 192), 2019 [3]	S	0.109	0.873	4.96	0.209	0.864	0.948	0.975
Monodepth2 (1024 × 320) 2019 [3]	S	0.107	0.849	4.764	0.201	0.874	0.953	0.977
Monodepth2 w/o pretraining, 2019 [3]	MS	0.127	1.031	5.266	0.221	0.836	0.943	0.974
Monodepth2 (640 × 192), 2019 [3]	MS	0.106	0.818	4.75	0.196	0.874	0.957	0.979
Monodepth2 (1024 × 320), 2019 [3]	MS	0.106	0.806	4.63	0.193	0.876	0.958	0.98
Ours w/o pretraining (640 × 192)	S	0.083	0.768	4.467	0.185	0.911	0.959	0.977
Ours (640 × 192)	S	0.080	0.747	4.346	0.181	0.918	0.961	0.978
Ours (1024 × 320)	S	0.077	0.723	4.233	0.179	0.922	0.961	0.978
Ours (1024 × 320) + PP	S	0.075	0.700	4.196	0.176	0.924	0.963	0.979