Skip to main content
EURASIP Journal on Audio, Speech, and Music Processing

Table 11 Influence of N iter parameter on the GMM gender recognition error rate

From: Evaluation of influence of spectral and prosodic features on GMM classification of Czech and Slovak emotional speech

Error rate (%)/Niter

100

200

400

600

800

1000

1200

1500

Minimum

1.75

2.56

2.41

2.45

2.44

2.46

2.46

2.46

Maximum

42.42

46.64

44.28

45.46

50

39.29

39.29

44.26

Mean

17.24

19.28

18.37

18.87

15.58

14.69

16.39

18.23

Back to article page