Intro; Abstract; Contents; 1 Introduction; 1.1 General Formulation of the Speech Enhancement Problem; 1.2 Organization of the Work; References; 2 Best Speech Enhancement Estimator in the Frequency Domain; 2.1 Signal Model and Problem Formulation; 2.2 Laws of Total Expectation and Total Variance; 2.3 Best Estimator; 2.4 Example with Gamma Distributions; 2.4.1 Reformulation of the Problem and Approximation; 2.4.2 Best Estimator; 2.5 A Brief Study of the Best Quadratic Estimator; 2.6 Generalization to the Multichannel Case; References; 3 Best Speech Enhancement Estimator in the Time Domain

3.1 Signal Model and Problem Formulation3.2 Best Estimator; 3.3 Best Linear Estimator; 3.4 Generalization to the Binaural Case; 3.4.1 Problem Formulation; 3.4.2 Best Estimator; 3.4.3 Best Widely Linear Estimator; References; 4 Speech Enhancement Via Correlation Coefficients; 4.1 Signal Model and Problem Formulation; 4.2 Linear Filtering and Correlation Coefficients; 4.3 Optimal Filters; 4.3.1 SPCC Between Filter Output and Desired Signal; 4.3.2 SPCC Between Filter Output and Noise Signal; 4.3.3 SPCC Between Filter Output and Filtered Desired Signal; 4.3.4 Other Possibilities; References

5 On the Output SNR in Speech Enhancement and Beamforming5.1 Signal Model and Problem Formulation; 5.2 Linear Filtering, Output and Fullmode Input SNRs; 5.3 Optimal Filters; 5.3.1 Rank-One Speech Covariance Matrix; 5.3.2 Rank-Deficient Speech Covariance Matrix; 5.3.3 Full-Rank Speech Covariance Matrix; 5.4 Application to Fixed and Superdirective Beamforming; References; 6 Speech Enhancement from the Fullband Output SNR Perspective; 6.1 Signal Model and Problem Formulation; 6.2 Speech Enhancement with Gains; 6.3 Determination of the Optimal Gains; 6.3.1 Maximization of the Fullband Output SNR

6.3.2 Minimization of the Fullband Output SNR6.4 Taking the Interframe Correlation Into Account; 6.5 Generalization to the Multichannel Case; References; Index