Skip to main content

Table 6 The accuracies (%) of models with different attention methods for each category on our LDWBC test set

From: Accurate classification of white blood cells by coupling pre-trained ResNet and DenseNet with SCAM mechanism

Attention method

B

M

E

N

L

No

97.62

71.63

100

99.30

98.23

CD

SE [51, 52]

100

85.58

98.95

99.67

96.86

 

ECA [53]

100

67.79

100

99.49

97.99

 

CAM [30]

100

78.37

98.95

99.49

98.48

SD

SAM [30]

97.62

84.13

98.95

99.30

97.45

TD

CAM // SAM [30]

97.62

77.40

98.95

99.39

97.99

 

CAM + SAM (CSAM) [30]

100

91.35

100

99.02

96.76

 

SAM + CAM (SCAM) [30]

95.24

90.38

100

99.44

96.86

  1. Best results are in bold; CD channel dimension, SD spatial dimension, TD two dimensions, // parallel, + sequential, B basophil, M monocyte, E eosinophil, N neutrophi, L lymphocyte
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us