Skip to main content

Table 3 Results of the baseline methods and the proposed method, for reverberant mixtures with the average over rooms A, B, C, and D in perceptual evaluation of speech quality (PESQ)

From: Localization based stereo speech source separation using probabilistic time-frequency masking and deep neural networks

Case

Methods

Room A

Room B

Room C

Room D

Mean

Determined

Mandel

2.34

2.07

2.34

1.96

2.18

 

Alinaghi

2.39

2.11

2.38

2.01

2.22

 

Propsoed

2.34

2.22

2.30

2.14

2.25

Underdetermined

Mandel

2.1

1.85

2.14

1.81

1.98

 

Alinaghi

2.15

1.87

2.18

1.84

2.01

 

Propsoed

2.06

1.97

2.19

1.85

2.02