An Improved Deep Residual Network-Based Semantic Simultaneous Localization and Mapping Method for Monocular Vision Robot

Table 4

The valuation of the two methods for SLAM tasks on various data sets.

Date set	Number of key frames		Number of map points		RMSE
	G-ORB	I-ORB	G-ORB	I-ORB	G-ORB	I-ORB
fre1_desk1	68	79	3157	4129	0.075	0.019
fre1_desk2	121	138	4236	5714	0.303	0.140
fre1_xyz	31	33	1714	1827	0.168	0.032