Table 6 Comparisons with other classic detection models on NEU-DET dataset.

From: A high precision YOLO model for surface defect detection based on PyConv and CISBA

Models

\(mAP^{val}_{50}\)

\(mAP^{val}_{50:95}\)

Improved-YOLOv566

73.08

37.57

Regularized YOLO67

80.77

47.62

MSFT-YOLO68

75.2

—

Improved Multi-Scale YOLO-v569

72

37.2

Kou’s70

72.2

—

EDNN65

72.4

—

YOLO-LFPD71

81.2

—

WFRE-YOLOv8s72

79.4

—

EPSC-YOLO (ours)

77.6

50

Back to article page