Skip to main content

Table 2 Average recognition results [%] of commonly used features subsets: acted speech (AS) and natural speech (NS)

From: Efficiency of chosen speech descriptors in relation to emotion recognition

Feature set

k-NN AS

k-NN NS

SVM AS

SVM NS

f 0

31.6

39.8

19.66

26.98

f 1f 3

38.8

39.5

16.88

27.5

Energy

30.4

47

29.95

46.87

LPC

36.7

57.8

29.91

60.51