Table 4.

The comparison results between our proposed model and other models on the TuSimple dataset.

Method	ACC (%)	PRE	REC	F1-Measure
ENet	97.84	0.854	0.952	0.900
SCNN	96.79	0.652	0.806	0.720
LaneNet	97.94	0.871	0.926	0.900
SegNet	97.73	0.809	0.869	0.838
SegNet_ConvLSTM	97.95	0.848	0.965	0.901
U-Net	97.91	0.861	0.945	0.902
U-Net_ConvLSTM	98.21	0.856	0.958	0.904
Proposed model	97.98	0.868	0.950	0.909