Skip to main content

Advertisement

Table 8 The MOS of the proposed algorithm and the contrast algorithm

From: Multi-resolution auditory cepstral coefficient and adaptive mask for speech enhancement with deep neural network

Noise type SNR (dB) MOS (score) Noise type SNR (dB) MOS (score)
Noisy speech Contrast1 Contrast2 Proposed algorithm Noisy speech Contrast1 Contrast2 Proposed algorithm
Buccaneer1 10 2.40 2.89 2.90 3.15 hfchannel 10 1.81 3.95 3.81 4.24
5 1.85 2.35 2.34 2.62 5 1.52 3.65 3.38 3.68
0 1.51 1.93 2.02 2.20 0 1.33 2.31 2.21 2.85
− 5 1.04 1.55 1.50 2.00 − 5 1.21 1.97 1.74 2.01
Buccaneer2 10 2.54 3.53 3.20 4.20 Leopard 10 2.12 3.54 3.41 3.75
5 1.62 2.86 2.83 3.27 5 1.86 2.83 2.64 3.12
0 1.34 2.34 2.32 2.56 0 1.80 2.13 2.21 2.74
− 5 0.51 1.91 1.92 2.17 − 5 1.73 2.05 2.01 2.58
Babble 10 2.23 4.02 4.13 4.36 m109 10 2.70 3.97 4.15 4.35
5 1.78 3.26 3.34 3.81 5 2.62 3.36 3.40 3.89
0 1.42 2.82 2.56 3.05 0 1.97 2.99 2.85 3.14
− 5 1.14 1.99 1.66 2.04 − 5 1.55 2.31 2.22 2.81
Destroyerengine 10 2.18 3.97 4.05 4.46 Machinegun 10 2.53 4.13 4.21 4.73
5 1.77 2.73 2.41 3.53 5 2.02 3.78 3.99 4.30
0 1.54 2.49 2.51 2.96 0 1.86 3.07 3.16 3.94
− 5 1.39 2.33 2.42 2.74 − 5 1.25 2.71 2.83 3.02
Destroyerops 10 2.50 3.96 3.82 4.12 Pink 10 2.52 3.69 3.50 3.89
5 2.02 3.70 3.16 3.82 5 2.11 2.38 2.43 2.58
0 1.60 2.35 2.82 3.38 0 1.97 2.09 2.15 2.50
− 5 1.24 2.92 2.57 2.97 − 5 1.65 1.81 1.92 2.32
f16 10 2.57 3.72 3.94 4.10 Volvo 10 2.30 3.64 3.83 4.26
5 2..01 3.22 3.48 3.82 5 1.96 2.84 2.93 3.58
0 1.80 3.01 3.15 3.45 0 1.65 2.29 2.37 3.00
− 5 1.30 2.95 2.62 2.98 − 5 1.22 1.93 1.81 2.75
Factory1 10 2.21 3.86 3.79 3.99 White 10 2.47 2.57 2.83 3.10
5 1.79 3.53 3.33 3.59 5 2.17 2.49 2.65 2.80
0 1.44 2.88 2.95 3.10 0 1.90 2.11 2.31 2.56
− 5 1.21 2.58 2.32 2.88 − 5 1.62 1.84 1.91 2.20
Factory2 10 2.53 3.84 3.91 4.39 Street 10 2.05 2.31 2.25 2.56
5 2.16 3.34 3.52 3.95 5 1.50 2.04 2.14 2.32
0 1.72 2.98 3.05 3.47 0 1.29 1.85 1.92 2.08
− 5 1.39 2.16 1.92 2.97 − 5 0.85 1.60 1.65 1.90
Office 10 2.01 2.96 3.04 3.11 Average 10 2.33 3.56 3.57 3.93
5 1.75 2.03 2.25 2.85 5 1.91 2.96 2.95 3.38
0 1.45 1.96 1.98 2.17 0 1.62 2.45 2.50 2.89
− 5 1.08 1.59 1.73 1.94 − 5 1.26 2.13 2.04 2.48