模型

Accuracy

/%

Precision

/%

Recall

/%

F1

/%

参数量

/M

FLOPs

/G

AlexNet

86.09

82.48

80.03

81.00

14.58

0.311

GoogleNet

85.09

81.67

82.04

81.42

5.97

1.59

VGG16

84.76

83.24

78.26

80.46

134.27

15.5

ResNet18

88.11

84.02

86.10

84.71

11.69

1.824

ResNet50

88.07

86.24

83.43

84.72

23.51

4.12

MobileNetV3

88.41

85.92

85.62

85.76

4.21

0.233

MobileNetV2

88.74

85.18

86.05

85.49

2.25

0.326

ShuffleNetV2

87.41

85.23

85.30

84.89

2.48

0.305

CIH-MobileNetV2

91.05

89.04

89.51

89.18

1.27

0.238