Table 5 Accuracy on the DB9 dataset.

From: Hand gesture recognition using sEMG signals with a multi-stream time-varying feature enhancement approach

Subject No.

Accuracy [%]

Subject No.

Accuracy [%]

Without

Channel Attention

With

Channel Attention

Without

Channel Attention

With

Channel Attention

1

99.84

100.00

40

95.19

100.00

2

89.26

97.50

41

80.53

100.00

3

99.69

100.00

42

82.01

97.50

4

94.88

97.05

43

80.19

100.00

5

98.48

97.50

44

90.87

100.00

6

99.14

97.50

45

89.17

97.50

7

97.46

97.50

46

92.54

97.50

8

99.49

97.50

47

92.88

100.00

9

100.00

97.50

48

84.09

97.50

10

99.49

97.50

49

92.08

100.00

11

99.96

100.00

50

82.50

100.00

12

99.30

100.00

51

88.30

97.50

13

99.73

97.50

52

90.80

100.00

14

95.94

100.00

53

93.22

100.00

15

99.02

100.00

54

93.83

100.00

16

100.00

100.00

55

86.70

97.50

17

99.84

95.00

56

92.84

97.50

18

99.22

100.00

57

95.95

100.00

19

95.94

100.00

58

87.23

100.00

20

97.89

100.00

59

91.02

100.00

21

99.65

100.00

60

92.23

100.00

22

98.05

100.00

61

94.51

97.50

23

99.88

95.00

62

88.14

100.00

24

99.84

97.50

63

86.02

97.50

25

99.57

100.00

64

85.68

100.00

26

99.73

100.00

65

85.11

100.00

27

99.84

99.75

66

91.25

100.00

28

99.84

100.00

67

90.53

97.50

29

87.46

100.00

68

98.83

100.00

30

93.48

100.00

69

97.07

97.50

31

91.35

100.00

70

92.73

100.00

32

99.28

97.5

71

88.13

100.00

33

88.98

100.00

72

90.82

100.00

34

90.57

100.00

73

90.20

100.00

35

89.62

97.50

74

98.12

100.00

36

88.07

97.50

75

87.34

100.00

37

95.98

100.00

76

87.15

100.00

38

86.48

100.00

77

98.44

100.00

39

92.80

100.00

Average

93.37

98.96