Efficient ConvNet Feature Extraction with Multiple RoI Pooling for Landmark-Based Visual Localization of Autonomous Vehicles

Table 4

Localization accuracy comparison of MRoI-FastRCNN-AlexNet versus FastRCNN-AlexNet and MRoI-FastRCNN-VGG-M1024 versus FastRCNN-VGG-M1024 in terms of maximum precision at 100% recall (Pr. at 100% Re.), maximum recall at 100% precision (Re. at 100% Pr.), and average precision (AP). The highest value with respect to each metric on each dataset is highlighted in bold. The middle italic values are the difference between MRoI-FastRCNN-AlexNet/VGG-M1024 and FastRCNN-AlexNet/VGG-M1024.

Method	UACampus			St. Lucia			Nordland			Mapillary
	Pr. at 100% Re.	Re. at 100% Pr.	AP	Pr. at 100% Re.	Re. at 100% Pr.	AP	Pr. at 100% Re.	Re. at 100% Pr.	AP	Pr. at 100% Re.	Re. at 100% Pr.	AP
MRoI-FastRCNN- AlexNet	98.9%	98.6%	99.7%	89.3%	68.2%	98.2%	75.2%	34.0%	92.1%	72.6%	31.2%	90.4%
	+4.0%	+12.7%	+0.5%	+5.3%	+7.0%	+1.4%	+6.2%	+10.1%	+3.7%	+5.4%	+15.2%	+4.2%
FastRCNN- AlexNet	94.9%	85.9%	99.2%	84.0%	61.2%	96.8%	69.0%	23.9%	88.4%	67.2%	16.0%	86.2%
MRoI-FastRCNN- VGG-M1024	98.3%	92.0%	99.6%	89.5%	74.4%	98.3%	75.0%	36.7%	92.3%	73.1%	24.3%	90.7%
	+2.8%	+7.5%	+0.3%	+5.5%	+26.0%	+1.9%	+6.2%	+1.4%	+3.6%	+6.9%	+10.2%	+5.6%
FastRCNN-VGG-M1024	95.5%	84.5%	99.3%	84.0%	48.4%	96.4%	68.8%	35.3%	88.7%	66.2%	14.1%	85.1%