Table 3 LAM-YOLO10n comparison experiment results.

From: Steel surface defect detection algorithm based on improved YOLOv10

Models

Precision(%)

Recall(%)

Map@0.5(%)

FPS(F/s)

SSD36

82.24

81.63

81.77

257

YOLOv5s37

85.67

83.02

84.27

236

Faster R-CNN38

87.11

85.23

86.55

274

YOLOv7-tiny39

91.89

89.74

89.86

207

YOLOv8n40

91.01

89.04

89.35

206

YOLOv11s41

93.83

91.36

92.35

168

YOLOv12n42

94.36

92.14

92.87

175

RT-DETR43

94.65

92.53

93.08

124

LAM-YOLOv10n

96.96

93.73

94.39

154

Back to article page