Table 5 The 3-fold comparison results.

From: Optimizing deep belief network for concrete crack detection via a modified design of ideal gas molecular dynamics

Method

Accuracy (%)

Specificity (%)

Fl Score

Precision (%)

Recall (%)

FCN9

84.176

75.948

67.015

81.442

75.196

YOLO10

74.797

81.741

73.285

84.38

77.119

CrackSegNet11

78.835

78.285

68.79

77.913

81.652

CNN12

82.029

77.297

74.966

89.164

88.258

CedNet13

81.554

76.574

78.851

85.109

85.654

DBN/MIGMM

88.408

92.15

80.901

93.051

91.961

Back to article page