Fast Vehicle and Pedestrian Detection Using Improved Mask R-CNN

Table 1

Frame comparison of R-CNN, Fast R-CNN, Faster R-CNN, and Mask R-CNN.

Network framework	R-CNN	Fast R-CNN	Faster R-CNN	Mask R-CNN
Propose time	2014	2015	2016	2017
Region proposal	Selective search	Selective search	RPN	RPN
Feature extraction	CNN	CNN + ROI pool	CNN + ROI pool	CNN + ROI align
Feature classification	SVM
Function	Classification, detection	Classification, detection	Classification, detection	Classification, detection, segmentation
Test time per image (with proposals)	47 seconds	2 seconds	0.2 seconds	0.2 seconds
mAP (VOC 2012)	62.4%	68.4%	70.4%	—