声明
ACKNOWLEDGEMENTS
Abstract
摘要
TABLE OF CONTENTS
LIST OF FIGURES
LIST OF TABLES
CHAPTER 1 INTRODUCTION
1.1 Speech Enhancement Algorithms
1.2 Noise Estimation Algorithms
1.3.Thesis Organization
CHAPTER 2 OVERVIEW OF SPECTRAL ENHANCEMENT TECHNIQUES
2.1 Spectral Subtraction
2.1.1 Spectral Subtraction Basic Principles
2.1.2 Drawbacks of Spectral Subtraction
2.2 Wiener Filtering
2.3 Soft Decision Method
2.4 Minimum Mean-Square Error(MMSE)
2.4.1 Optimally Modified Log-Spectral Amplitude Estimator(OM-LSA)
CHAPTER 3 OVERVIEW OF EXISTING NOISE ESTIMATION ALGORITHMS
3.1 Minimum Statistics Noise Estimation Algorithm
3.1.1 Principles of the Minimum Statistics Method
3.1.2 Performance evaluation
3.1.3 Limitations of the Minimum Statistics Method
3.2 Minima Controlled Recursive Averaging(MCRA)
3.2.1 Overview of the MCRA technique
3.2.2 Performance Evaluation
3.2.3 Limitations of the MCRA technique
3.3 Improved MCRA Algorithm based on Conditional MAP
3.4 Continuous Spectral Minimum Tracking
3.5 MCRA-2 Algorithm
CHAPTER 4 PROPOSED ALGORITHMS FOR ENHANCEMENT OF NOISY SPEECH IN ADVERSE ENVIRONMENTS
4.1 Proposed Speech Enhancement Method Based MMSE
4.1.1 Proposed Gain Modification
4.1.2 The Proposed A Priori SNR Estimation
4.1.3 The Proposed A Priori Speech Absence Estimation Method
4.1.4 Performance Evaluation
4.2.Proposed Noise Estimation Algorithm
4.2.1 Description of the Proposed Noise Estimation Algorithm
4.2.2 Performance Evaluation
SUMMARY AND CONCLUSIONS
REFERENCES
APPENDIX A SCHOLARLY PUBLICATION(S)
APPENDIX B JOINED SCIENTIFIC RESEARCH PROJECTS DURING MASTER’S STUDIES