notesum.ai

Published at November 29

MonoPP: Metric-Scaled Self-Supervised Monocular Depth Estimation by Planar-Parallax Geometry in Automotive Applications

cs.CV

cs.AI

cs.LG

cs.RO

Released Date: November 29, 2024

Authors: Gasser Elazab¹, Torben Gräber, Michael Unterreiner¹, Olaf Hellwich²

Aff.: ¹CARIAD SE; ²Technische Universität Berlin

Arxiv: http://arxiv.org/pdf/2411.19717v1

Refer to caption

	Year	Method	Train	Abs Rel $\downarrow$	Sq Rel $\downarrow$	RMSE $\downarrow$	RMSE log $\downarrow$	$\delta<1.25$ $\uparrow$	$\delta<1.25^{2}$ $\uparrow$	$\delta<1.25^{3}$ $\uparrow$
scaled by GT	2019	Monodepth2 [14]	M	0.115	0.903	4.863	0.193	0.877	0.959	0.981
	2020	PackNet-SFM [16]	M	0.111	0.785	4.601	0.189	0.878	0.960	0.982
	2021	CADepth [48]	M	0.110	0.812	4.686	0.187	0.882	0.961	0.981
	2022	VADepth [44]	M	0.104	0.774	4.552	0.181	0.892	0.965	0.983
	2022	MonoFormer [3]	M	0.108	0.806	4.594	0.184	0.884	0.963	0.983
	2022	MonoViT [56]	M	0.099	0.708	4.372	0.175	0.900	0.967	0.984
	2023	Lite-Mono [53]	M	0.107	0.765	4.561	0.183	0.886	0.963	0.983
	2023	Lite-Mono-S [53]	M	0.110	0.802	4.671	0.186	0.879	0.961	0.982
	2023	TriDepth [7]	M	0.093	0.665	4.272	0.172	0.907	0.967	0.984
	MonoPP (ours)		M	0.105	0.776	4.640	0.185	0.891	0.962	0.982
w/o scaling	2019	Monodepth2** [14]	M+camH	0.126	0.973	4.880	0.198	0.864	0.957	0.980
	2020	DNet [47]	M+camH	0.118	0.925	4.918	0.199	0.862	0.953	0.979
	2020	Zhao et al. [55]	M+SC	0.146	1.084	5.445	0.221	0.807	0.936	0.976
	2020	PackNet [16]	M+V	0.111	0.829	4.788	0.199	0.864	0.954	0.980
	2021	Wagstaff et al. [39]	M+Pose	0.123	0.996	5.253	0.213	0.840	0.947	0.978
	2021	Wagstaff et al. [39]	M+camH	0.155	1.657	5.615	0.236	0.809	0.924	0.959
	2021	Sui et al. [36]	M+camH	0.128	0.936	5.063	0.214	0.847	0.951	0.978
	2022	VADepth [44]	M+camH	0.109	0.785	4.624	0.190	0.875	0.960	0.982
	2022	DynaDepth [54]	M+Pose	0.109	0.787	4.705	0.195	0.869	0.958	0.981
	2023	Lee et al. [25]	M+Pose	0.141	1.117	5.435	0.223	0.804	0.942	0.977
	2024	FUMET [22]	M+SI	0.108	0.785	4.736	0.195	0.871	0.958	0.981
	MonoPP (ours)		M+camH	0.107	0.835	4.658	0.186	0.891	0.962	0.982