Skip to main content

Table 4 PESQ values for the W β-SA estimator with different schemes of parameter β, case of white noise

From: Bayesian STSA estimation using masking properties and generalized Gamma prior for speech enhancement

Input SNR (dB) −10 −5 0 5 10
Input noisy speech 1.13 1.26 1.47 1.75 2.06
Choice of β = 1.82 1.48 1.69 2.00 2.32 2.68
Choice of β by (11) 1.53 1.74 2.08 2.39 2.72
Choice of β by (12) 1.52 1.74 2.06 2.42 2.75
Proposed choice of β 1.54 1.77 2.14 2.49 2.81