Table 2 Comparison of classic network performance.

From: Finger vein recognition based on bilinear fusion of multiscale features

Method

Accuracy/%

Time/ms

Parameter/M

Flops/GFlops

FV-USM

SDUMLA

VGG-1619

95.83

95.71

49

44.0

3.20

DesNet20

96.95

98.16

34

26.7

0.17

AlexNet21

92.28

96.32

16

19.7

0.43

ResNet-1822

96.03

97.79

16

11.0

2.40

MSFBF-Net

99.59

99.53

3

8.2

0.21

Back to article page