Fig. 8From: Bayesian STSA estimation using masking properties and generalized Gamma prior for speech enhancementSegmental SNR versus global SNR for different W β-SA estimators, (a) white noise, (b) babble noise, and (c) car noiseBack to article page