Skip to main content

Table 6 Simulation 5: 5 A, 5 B, and 5 C are the SIA for crowded talking NSN and G.712 type handset at 16 kHz for different signal to noise ratio levels for the TIMIT, SITW, and NIST 2008, respectively, at mixture size 256

From: Evaluation of a speaker identification system with and without fusion using three databases in the presence of noise and handset effects

Methods

SNR0 dB

SNR5 dB

SNR10 dB

SNR15 dB

SNR20 dB

SNR25 dB

SNR30 dB

Simulation 5 A: the SIA for crowded talking NSN and G.712 type handset at 16 kHz for

FWMFCC (f 1)

9.17%

18.33%

35%

50.83%

66.67%

74.17%

80%

CMVNMFCC (f 2)

7.5%

19.17%

34.17%

55.83%

69.17%

81.67%

87.5%

FWPNCC (g 1)

1.67%

2.5%

15.83%

29.17%

43.33%

56.67%

59.17%

CMVNPNCC (g 2)

1.67%

5%

19.17%

35%

54.17%

60.83%

68.33%

Fusion decision

(f 1- g 2)

(f 2- g 2)

(f 1- g 2)

(f 2- g 2)

(f 2- g 2)

(f 2- g 2)

(f 2- g 2)

Fused ω 1 =0.9

10%

19.17%

35.83%

57.5%

70.83%

83.33%

87.5%

Fused ω 2 =0.8

10%

16.67%

36.67%

59.17%

71.67%

83.33%

Fused ω 3 =0.77

10%

16.67%

36.67%

60%

72.5%

83.33%

88.33%

Fused ω 4 =0.7

8.33%

16.67%

37.5%

61.67%

74.17%

84.17%

88.33%

Fusion max

2.5%

9.17%

39.17%

52.5%

73.33%

84.17%

88.33%

Fusion mean

5%

15%

38.33%

62.5%

73.33%

82.5%

89.17%

Simulation 5 B: the SIA for crowded talking NSN and G.712 type handset at 16 kHz for

FWMFCC (f 1)

18.33%

33.33%

45.83%

64.17%

73.33%

75.83%

78.33%

CMVNMFCC (f 2)

15.83%

30%

43.33%

59.17%

72.5%

75.83%

77.5%

FWPNCC (g 1)

5%

15%

33.33%

59.17%

71.67%

76.67%

79.17%

CMVNPNCC (g 2)

4.17%

12.5%

30%

53.33%

70%

75.83%

80.83%

Fusion decision

(f 1- g 1)

(f 1- g 1)

(f 1- g 1)

(f 1- g 1)

(f 1- g 1)

(f 1- g 1)

(f 1- g 2)

Fused ω 1 =0.9

20%

65%

48.33%

67.5%

73.33%

75.83%

80%

Fused ω 2 =0.8

18.33%

61.67%

50%

68.33%

73.33%

75.83%

80%

Fused ω 3 =0.77

17.5%

60%

50.83%

69.17%

73.33%

75.83%

80%

Fused ω 4 =0.7

17.5%

57.5%

53.33%

70%

73.33%

77.5%

80%

Fusion max

14.17%

48.33%

46.67%

65.83%

73.33%

76.67%

80.83%

Fusion mean

11.67%

45%

50.83%

72.5%

75%

78.33%

Simulation 5 C: the SIA for crowded talking NSN and G.712 type handset at 16 kHz for

FWMFCC (f 1)

7.5%

12.5%

24.17%

30%

37.5%

47.5%

66.67%

CMVNMFCC (f 2)

3.33%

10.83%

18.33%

28.33%

40.83%

46.67%

67.5%

FWPNCC (g 1)

3.33%

11.67%

29.17%

44.17%

67.5%

78.33%

80.83%

CMVNPNCC (g 2)

2.5%

10%

24.17%

45%

68.33%

79.17%

82.5%

Fusion decision

(f 1- g 1)

(f 1- g 1)

(f 1- g 1)

(f 1- g 2)

(f 2- g 2)

(f 1- g 2)

(f 2- g 2)

Fused ω 1 =0.9

6.67%

15%

24.17%

34.17%

45.83%

55.83%

70.83%

Fused ω 2 =0.8

10%

15%

24.17%

35%

48.33%

60.83%

75.83%

Fused ω 3 =0.77

10%

15%

25.83%

36.67%

49.17%

61.67%

77.5%

Fused ω 4 = 0.7

10%

15%

28.33%

40.83%

49.17%

64.17%

80%

Fusion max

8.33%

15.83%

29.17%

45.83%

51.67%

70%

77.5%

Fusion mean

8.33%

17.5%

30%

45%

57.5%

73.33%

  1. The colored data reflected three different databases and the highest SIA for each database: red for TIMIT, blue for SITW and Violet for NIST 2008 database. The colored italic entries represent the highest SIA