Format:
1 Online-Ressource (xii, 70 Seiten)
,
Illustrationen
Edition:
Electronic reproduction; Available via World Wide Web
ISBN:
9781627051446
Series Statement:
Synthesis lectures on speech and audio processing #11
Content:
As speech processing devices like mobile phones, voice controlled devices, and hearing aids have increased in popularity, people expect them to work anywhere and at any time without user intervention. However, the presence of acoustical disturbances limits the use of these applications, degrades their performance, or causes the user difficulties in understanding the conversation or appreciating the device. A common way to reduce the effects of such disturbances is through the use of single-microphone noise reduction algorithms for speech enhancement. The field of single-microphone noise reduction for speech enhancement comprises a history of more than 30 years of research. In this survey, we wish to demonstrate the significant advances that have been made during the last decade in the field of discrete Fourier transform domain-based single-channel noise reduction for speech enhancement. Furthermore, our goal is to provide a concise description of a state-of-the-art speech enhancement system, and demonstrate the relative importance of the various building blocks of such a system. This allows the non-expert DSP practitioner to judge the relevance of each building block and to implement a close-to-optimal enhancement system for the particular application at hand
Content:
3. DFT-based speech enhancement methods-signal model and notation --
Content:
2. Single channel speech enhancement-general principles -- 2.1 Analysis-modification-synthesis (AMS) system -- 2.2 Finding the target estimate -- 2.3 A priori knowledge and assumptions -- 2.3.1 Taking speech signal characteristics into account -- 2.3.2 Taking noise process characteristics into account -- 2.3.3 Taking the human auditory system into account --
Content:
10. Future directions -- References -- Authors' biographies
Content:
4. Speech DFT estimators -- 4.1 Statistical modeling assumptions -- 4.2 Spectral subtraction -- 4.3 Linear MMSE estimators -- 4.4 Non-linear MMSE estimators --
Content:
5. Speech presence probability estimation -- 5.1 A posteriori speech presence probability -- 5.2 Estimation of the model parameter H1 -- 5.2.1 Short-term adaptive estimate -- 5.2.2 Fixed optimal H1 -- 5.3 Choosing the prior probabilities -- 5.3.1 Adaptive prior probabilities -- 5.3.2 Fixed prior probabilities -- 5.4 Avoiding outliers --
Content:
6. Noise PSD estimation -- 6.1 Methods based on VAD -- 6.2 Methods based on minimum power level tracking -- 6.3 SPP-based noise PSD estimation -- 6.4 MMSE-based estimation of the noise PSD -- 6.5 DFT-subspace estimation of the noise PSD --
Content:
7. Speech PSD estimation -- 7.1 Maximum likelihood estimation and decision-directed approach -- 7.2 Kalman-type filtering, Garch modeling, and noncausal estimation -- 7.3 Temporal cepstrum smoothing -- 7.4 Comparison of the estimators --
Content:
8. Performance evaluation methods -- 8.1 Evaluating quality aspects of enhanced speech -- 8.1.1 Listening tests -- 8.1.2 Instrumental test methods -- 8.2 Evaluating intelligibility of enhanced speech -- 8.2.1 Listening tests -- 8.2.2 Instrumental test methods --
Content:
9. Simulation experiments with single-channel enhancement systems --
Content:
Acknowledgments -- Glossary -- 1. Introduction --
Note:
Includes bibliographical references
,
Acknowledgments; Glossary; Introduction; Single Channel Speech Enhancement-General Principles; Analysis-Modification-Synthesis (AMS) System; Finding the Target Estimate; A priori Knowledge and Assumptions; Taking Speech Signal Characteristics into Account; Taking Noise Process Characteristics into Account; Taking the Human Auditory System into Account; DFT-Based Speech Enhancement Methods-Signal Model and Notation; Speech DFT Estimators; Statistical Modeling Assumptions; Spectral Subtraction; Linear MMSE Estimators; Non-linear MMSE Estimators; Speech Presence Probability Estimation
,
A posteriori Speech Presence ProbabilityEstimation of the Model Parameter H1; Short-term Adaptive Estimate; Fixed Optimal H1; Choosing the Prior Probabilities; Adaptive Prior Probabilities; Fixed Prior Probabilities; Avoiding Outliers; Noise PSD Estimation; Methods Based on VAD; Methods Based on Minimum Power Level Tracking; SPP-Based Noise PSD Estimation; MMSE-Based Estimation of the Noise PSD; DFT-subspace Estimation of the Noise PSD; Speech PSD Estimation; Maximum Likelihood Estimation and Decision-directed Approach; Kalman-type Filtering, GARCH Modeling, and Noncausal Estimation
,
Temporal Cepstrum SmoothingComparison of the Estimators; Performance Evaluation Methods; Evaluating Quality Aspects of Enhanced Speech; Listening Tests; Instrumental Test Methods; Evaluating Intelligibility of Enhanced Speech; Listening Tests; Instrumental Test Methods; Simulation Experiments with Single-Channel Enhancement Systems; Future Directions; References; Authors' Biographies
,
Electronic reproduction; Available via World Wide Web
,
Mode of access: World Wide Web.
,
System requirements: Adobe Acrobat Reader.
Additional Edition:
ISBN 9781627051439
Additional Edition:
Print version Hendriks, Richard C. DFT-domain based single-microphone noise reduction for speech enhancement [Williston, VT] : Morgan & Claypool, 2013 ISBN 9781627051439
Language:
English
DOI:
10.2200/S00473ED1V01Y201301SAP011
Bookmarklink