Skip to main content

Table 5 Confusion matrix (HTI 2014–2015 dataset)

From: Laban movement analysis and hidden Markov models for dynamic 3D gesture recognition

 

A

B

C

D

E

F

G

H

I

J

K

A

56..6

0

2.4

2.6

0.1

3.5

33.8

0

0.1

0.9

0

B

21.7

51.4

0.4

4.5

0.1

0.1

0.9

0

0.1

20.8

0

C

38.2

0

40.5

5.1

1.1

0.4

12.4

0

0

0.2

2.1

D

42..4

1.7

5.2

40.8

0.2

1

2.8

0.1

0.2

5.6

0

E

29..2

0.1

12.4

3.2

23.7

5

12

0.5

7.9

0.8

5.2

F

20.4

0.3

2.2

2.5

4.8

60.5

4.7

0

3.8

0.6

0.2

G

32.3

0.5

6.4

5.6

7.1

2.6

41.2

0

0.7

0.7

2.9

H

18.1

0.4

0.2

3.4

1.9

1.4

1.8

59.9

11.6

0.5

0.8

I

21.8

0.6

0.4

3.7

1.3

1.8

1.7

39.6

28.1

0.8

0.2

J

17.1

14.9

0.6

2.6

1.8

0.5

1.3

0

0

61.1

0.1

K

16.4

0.1

1.6

4.3

5

1.2

3.5

3.2

4.3

0.7

59.7

  1. In italics: the best recognition scores for each row (category). The (i, j) entry in this matrix represents the percentage of frames of a sequence in the i category that have been recognized as frames from the j category