Table 4 Presents a comparison of classification data for classical models on the NEU dataset.

From: High precision classification of hot rolled strip steel surface defects using dual path features and entropy attention fusion

Model

Accuracy(%)

Precision(%)

F1-score(%)

Recall(%)

ResNet-101

99.23

98.91

99.02

98.86

Yolov10

96.35

95.47

96.23

95.79

VGG16

91.24

90.13

91.72

91.97

ResNet18

93.25

93.53

92.21

93.72

DenseNet121

96.39

96.77

96.39

96.36

Ourwork

99.17

99.19

99.17

99.16

Back to article page