Table 4 Ablation experiment results of PCE-YOLO

From: An efficient algorithm for pedestrian fall detection in various image degradation scenarios based on YOLOv8n

Model

CPA-Enhancer

Pm-C2f

Inner-EIoU

mAP/%

Params/M

FLOPs/G

FPS

M1

\(\times\)

\(\times\)

\(\times\)

85.52

3.15

8.2

273.5

M2

\(\checkmark\)

\(\times\)

\(\times\)

88.65

3.51

21.0

196.4

M3

\(\times\)

\(\checkmark\)

\(\times\)

86.01

2.19

5.8

288

M4

\(\checkmark\)

\(\checkmark\)

\(\times\)

88.95

2.8

18.6

210.5

M5

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

90.04

2.8

18.6

210.5

Back to article page