You are previewing Speech Enhancement.
O'Reilly logo
Speech Enhancement

Book Description

Speech enhancement is a classical problem in signal processing, yet still largely unsolved. Two of the conventional approaches for solving this problem are linear filtering, like the classical Wiener filter, and subspace methods. These approaches have traditionally been treated as different classes of methods and have been introduced in somewhat different contexts. Linear filtering methods originate in stochastic processes, while subspace methods have largely been based on developments in numerical linear algebra and matrix approximation theory.

This book bridges the gap between these two classes of methods by showing how the ideas behind subspace methods can be incorporated into traditional linear filtering. In the context of subspace methods, the enhancement problem can then be seen as a classical linear filter design problem. This means that various solutions can more easily be compared and their performance bounded and assessed in terms of noise reduction and speech distortion. The book shows how various filter designs can be obtained in this framework, including the maximum SNR, Wiener, LCMV, and MVDR filters, and how these can be applied in various contexts, like in single-channel and multichannel speech enhancement, and in both the time and frequency domains.




    • First short book treating subspace approaches in a unified way for time and frequency domains, single-channel, multichannel, as well as binaural, speech enhancement.
    • Bridges the gap between optimal filtering methods and subspace approaches.
    • Includes original presentation of subspace methods from different perspectives.

Table of Contents

  1. Cover image
  2. Title page
  3. Table of Contents
  4. Copyright
  5. Chapter 1. Introduction
    1. Abstract
    2. 1.1 History and Applications of Subspace Methods
    3. 1.2 Speech Enhancement from a Signal Subspace Perspective
    4. 1.3 Scope and Organization of the Work
    5. References
  6. Chapter 2. General Concept with the Diagonalization of the Speech Correlation Matrix
    1. Abstract
    2. 2.1 Signal Model and Problem Formulation
    3. 2.2 Linear Filtering with a Rectangular Matrix
    4. 2.3 Performance Measures
    5. 2.4 Optimal Rectangular Filtering Matrices
    6. References
  7. Chapter 3. General Concept with the Joint Diagonalization of the Speech and Noise Correlation Matrices
    1. Abstract
    2. 3.1 Signal Model and Problem Formulation
    3. 3.2 Linear Filtering with a Rectangular Matrix
    4. 3.3 Performance Measures
    5. 3.4 Optimal Rectangular Filtering Matrices
    6. 3.5 Another Signal Model
    7. References
  8. Chapter 4. Single-Channel Speech Enhancement in the Time Domain
    1. Abstract
    2. 4.1 Signal Model and Problem Formulation
    3. 4.2 Linear Filtering with a Rectangular Matrix
    4. 4.3 Performance Measures
    5. 4.4 Optimal Rectangular Filtering Matrices
    6. 4.5 Single-Channel Noise Reduction Revisited
    7. References
  9. Chapter 5. Multichannel Speech Enhancement in the Time Domain
    1. Abstract
    2. 5.1 Signal Model and Problem Formulation
    3. 5.2 Linear Filtering with a Rectangular Matrix
    4. 5.3 Performance Measures
    5. 5.4 Optimal Rectangular Filtering Matrices
    6. References
  10. Chapter 6. Multichannel Speech Enhancement in the Frequency Domain
    1. Abstract
    2. 6.1 Signal Model and Problem Formulation
    3. 6.2 Linear Array Model
    4. 6.3 Performance Measures
    5. 6.4 Optimal Filters
    6. References
  11. Chapter 7. A Bayesian Approach to the Speech Subspace Estimation
    1. Abstract
    2. 7.1 Signal Model and Problem Formulation
    3. 7.2 Estimation Based on the Minimum Mean-Square Distance
    4. 7.3 A Closed-Form Solution Based on the Bingham Posterior
    5. References
  12. Chapter 8. Evaluation of the Time-Domain Speech Enhancement Filters
    1. Abstract
    2. 8.1 Evaluation of Single-Channel Filters
    3. 8.2 Evaluation of Multichannel Filters
    4. References
  13. Index