You are viewing the site in preview mode

Skip to main content

Table 4 Training time and prediction time for ML-DRSNet at different window sizes and step sizes

From: A multi-label deep residual shrinkage network for high-density surface electromyography decomposition in real-time

Window size

(data point)

Step size

(data point)

Data size

(sample)

Training time

(s/epoch)

Prediciton time

(ms/sample)

20

10

47,265

427.53

5.45

20

23,632

215.45

5.50

30

15,755

144.92

5.45

40

11,816

1096.91

5.43

50

9452

88.59

5.42

60

10

47,261

435.06

5.37

20

23,630

218.68

5.47

30

15,754

145.00

5.39

40

11,815

113.74

5.73

50

9452

91.68

5.50

100

10

47,257

432.02

5.36

20

23,628

217.37

5.34

30

15,752

146.28

5.34

40

11,814

111.43

5.29

50

9452

87.43

5.22

140

10

47,253

421.53

5.21

20

23,626

212.15

5.25

30

15,751

146.88

5.40

40

11,813

112.11

5.29

50

9450

90.82

5.29

Average

198.43 ± 126.19

5.39 ± 0.12

  1. Training time and prediction time for ML-DRSNet at different window sizes and step sizes using data from subject S10 (10% MVC) for reference