Table 4 Comparison experiment results on LLAMAS.

From: Aggregate global features into separable hierarchical lane detection transformer

Method

F1 (%)

Precision (%)

Recall (%)

SCNN34

94.25

94.11

94.39

RESA35

94.00

93.52

94.47

LaneATT43

93.46

96.92

90.24

PolyLaneNet41

88.40

88.87

87.93

Ours

94.37

94.21

94.53

  1. Bold, italic, and bolditalic indicate the first, second, and third scores, respectively.
Back to article page