A Family of Bayesian STSA Estimators for the Enhancement of Speech With Correlated Frequency Components

Free registration required

Executive Summary

In Bayesian Short-Time Spectral Amplitude (STSA) estimation for single-channel speech enhancement, the spectral components are traditionally assumed uncorrelated. However, this assumption is inexact since some correlation is present in practice. The authors, thus investigate a multi-dimensional STSA estimator that assumes correlated frequency components. Since, the closed-form solution of this optimum estimator is not readily available, they previously derived closed-form expressions for an upper and a lower bound on the desired estimator. In this paper, they study the proximity between the upper and the lower bounds.

  • Format: PDF
  • Size: 112.46 KB