Intelligent Semantic Segmentation for Self-Driving Vehicles Using Deep Learning

Table 3

Comparison between proposed model and other baseline models.

Model	mCA (%)	mIoU (%)	#Params (m)
SegNet-Basic [19]	62.9	46.2	—
SegNet [18]	65.2	55.6	29.5
FCN-8s [17]	—	57	134.5
ApesNet [24]	69.3	48	—
ENet [25]	68.3	51.3	0.36
ESPNet [26]	68.3	55.6	0.36
ESCNet [27]	70.9	56.1	0.185
DeepLab-LFOV [23]	—	61.6	37.3
Dilated-8 [22]	—	65.3	140.8
Proposed model	78.03	58.275	5.2