Table 2 The performance comparison of the basic network model

ResNet50

T

0.821

[0.802 0.840]

78.4

[76.4 80.4]

83.5

[80.2 86.8]

65.6

[61.3 69.9]

48.7

[45.6 51.8]

97.5

[96.4 98.6]

V

0.699

[0.667 0.731]

67.7

[63.4 72.0]

72.3

[64.1 80.5]

61.5

[57.6 65.4]

43.5

[39.9 47.1]

81.7

[80.2 83.2]

I-T

0.676

[0.656 0.697]

63.5

[60.9 66.1]

71.5

[63.3 79.7]

62.3

[55.6 69.0]

39.2

[37.7 40.7]

80.6

[76.8 84.5]

ResNet101

T

0.763

[0.749 0.777]

74.6

[73.6 75.6]

78.6

[74.6 82.6]

72.9

[71.3 74.5]

65.7

[62.5 68.9]

89.0

[83.9 94.1]

V

0.649

[0.611 0.668]

62.0

[56.8 67.2 ]

71.6

[61.4 81.8]

59.7

[58.3 61.1]

44.3

[42.0 46.6]

87.8

[83.4 92.3]

I-T

0.637

[0.604 0.670]

61.4

[55.3 67.5]

71.4

[63.2 80.1]

56.5

[53.0 60.0]

42.5

[37.3 47.7]

82.6

[77.1 88.1]

DenseNet121

T

0.880

[0.866 0.892]

84.5

[83.2 85.8]

92.6

[90.3 94.9]

72.1

[69.3 74.9]

56.9

[53.6 60.2]

98.1

[97.7 98.5]

V

0.798

[0.765 0.831]

71.2

[66.5 75.9]

93.0

[91.7 94.3]

59.3

[52.2 66.4]

51.3

[44.6 58.0]

90.2

[85.7 94.7]

I-T

0.792

[0.777 0.807]

68.1

[65.9 70.3]

94.5

[93.4 95.6]

60.3

[56.3 64.3]

48.4

[45.6 51.2]

89.8

[87.5 92.1]

VGG19

T

0.771

[0.758 0.784]

72.8

[67.3 77.7]

87.3

[82.3 92.3]

69.9

[61.0 76.8]

60.8

[54.6 67.0]

91.3

[90.1 92.5]

V

0.683

[0.664 0.702]

63.5

[61.0 66.0]

79.5

[75.2 83.8]

64.2

[60.0 68.4]

47.5

[46.2 48.8]

86.1

[83.5 88.7]

I-T

0.679

[0.651 0.707]

62.2

[60.2 64.2]

72.4

[67.3 77.5]

58.1

[55.3 60.9]

45.2

[43.8 46.6]

83.6

[80.6 86.6]

Inception V3

T

0.826

[0.815 0.837]

80.5

[79.3 81.7]

88.5

[84.7 92.3]

73.1

[70.8 75.4]

54.2

[52.6 55.8]

93.4

[91.6 95.2]

V

0.706

[0.683 0.729]

70.8

[66.7 74.9]

79.3

[76.1 82.5]

64.5

[58.5 70.5]

41.3

[37.9 44.7]

85.6

[82.7 88.5]

I-T

0.695

[0.657 0.733]

69.3

[66.5 72.1]

75.1

[71.6 78.6]

61.3

[54.5 68.1]

40.8

[38.0 43.6]

82.3

[79.1 85.5]