Table 25.

Comparison of the Results on the KITTI, NCLT, and KAIST Datasets.

KITTI Dataset						NCLT Dataset	KAIST Dataset
Task: 3DPR						Task: 3DPR
	Accuracy			Ref	Recall @ 1 %	Recall @ 1 %	Recall @ 1 %
Ref	Descriptors	Pair of Matching Segments	Candidate Matching	[88]	98.1	-	95.2
[86]	Siamese	80%	30%	[84]	96.9	97.0	-
	Group-based	-	50%	-
	Constructrive	-	50%