Skip to main content

Table 1 SNR results

From: A comparison of several computational auditory scene analysis (CASA) techniques for monaural speech segregation

 

N0

N1

N2

N3

N4

N5

N6

N7

N8

N9

Average

Mixture

−3.26

−4.07

10.19

4.34

3.99

−5.82

1.90

6.62

10.37

0.73

2.49

Hu and Wang [11]

16.34

7.83

16.71

8.32

10.88

14.41

16.89

11.97

14.44

5.27

12.30

Hu and Wang [12]

24.50

13.50

20.30

13.40

11.99

22.40

18.60

15.11

17.60

8.66

16.60

Zhang and Liu [13]

17.07

5.94

17.26

6.26

8.50

15.18

16.23

11.50

14.43

7.40

11.97

Zhang and Liu [14]

17.86

8.16

18.27

8.26

11.28

16.04

17.46

11.93

14.84

4.98

12.90

Li and Guan [17]

11.13

3.50

14.41

5.21

6.66

12.93

14.66

9.39

11.50

3.96

9.33

True pitch

16.33

8.35

17.71

8.79

11.56

15.06

17.76

12.31

15.32

6.04

12.92

Narrow band [23]

9.88

6.74

11.44

6.94

8.95

8.33

11.31

9.15

10.60

3.98

8.73

Comb filter

3.12

3.01

13.28

8.72

8.32

2.25

6.56

10.57

13.19

5.39

7.44

Wang–Brown [24]

11.31

4.93

11.19

5.65

8.72

10.44

11.15

9.22

10.84

2.66

8.61

Spectral subtraction

18.35

3.05

16.00

6.14

8.32

−5.51

4.85

8.23

10.90

2.46

7.27

Ideal binary mask

20.76

9.04

22.90

9.72

13.19

18.40

21.53

15.78

18.10

10.5

15.99