Skip to main content

Table 2 YOLOv8n-modify comparative analysis with alternative methods

From: Applying optimized YOLOv8 for heritage conservation: enhanced object detection in Jiangnan traditional private gardens

Modules

Adopted modules

Precision (%)

Recall(%)

mAP@0.5(%)

mAP@0.5:0.95(%)

FPS(%)

YOLOv3n

None

55

48.6

53.1

25.8

27.2

YOLOv5n

None

57.4

51

51

22.4

38.3

YOLOv8n

None

57.4

53.5

55.3

26.1

38.9

YOLOv8n-modify

DBB + BiFPN + DyHead Attention

66.1

46.7

57.1

29.2

14.8