首页--工业技术论文--无线电电子学、电信技术论文--通信论文--电声技术和语音信号处理论文--语音信号处理论文

语音信号中敲击按键声音的检测和抑制

摘要第5-6页
Abstract第6-7页
List of symbols and abbreviations第17-18页
Chapter 1 Introduction第18-34页
    1.1. Background and significance of the noise suppression research in speechsignals第18-22页
    1.2. Overview第22-30页
    1.3. Main tasks第30-31页
    1.4. Dissertation organization第31-34页
Chapter 2 Speech Enhancement Models第34-56页
    2.1. Speech Enhancement第34-36页
        2.1.1. Single Channel Speech Enhancement第34-35页
        2.1.2. Multichannel Speech Enhancement第35-36页
    2.2. Spectral Subtraction第36-40页
        2.2.1. The Basic Spectral Subtraction Method第37-38页
        2.2.2. Spectral Subtraction as a Filter第38-39页
        2.2.3. Effect of Windows on the Spectral Subtraction第39页
        2.2.4. Spectral Subtraction Technique Drawbacks第39-40页
    2.3. Transient Interference Suppression in Speech Signals based on theOM-LSA Algorithm第40-44页
        2.3.1. Transient Event Power Spectral Density Estimation第41-43页
        2.3.2. Enhancement of Speech第43-44页
    2.4. Speech Production Fundamentals第44-45页
    2.5. Analysis of Noise第45-46页
    2.6. Time-Frequency Representation第46-49页
    2.7. Short-Time Fourier Transform(STFT)第49-51页
    2.8. Keystroke Noise Model第51-54页
        2.8.1. Keystrokes Analysis第51-52页
        2.8.2. Keystrokes Effects on Speech Signals第52-54页
    2.9. Summary第54-56页
Chapter 3 Non-negative Matrix Factorization第56-66页
    3.1. Similarities to other Matrix Factorizations Techniques第57页
    3.2. Need of Non-Negativity第57-58页
    3.3. NMF Review Papers第58页
    3.4. Defining a Cost Function第58-60页
        3.4.1. Maximum Likelihood Estimate第58-59页
        3.4.2. Sparsity Enforcement第59-60页
    3.5. Costfunction Minimization第60-63页
        3.5.1. Derivation of the Update Rules for Sparse Costfunctions第60-63页
    3.6. Complete SNMF Algorithm第63-64页
        3.6.1. Sparse Multiplicative Update Rules for Least Squares NMF第63页
        3.6.2. Multiplicative Update Rules for Kullback-Leibler SNMF第63-64页
    3.7. Convergence第64页
    3.8. Summary第64-66页
Chapter 4 Two Stage Approaches for the Detection and Suppression of typedKeystrokes in Speech Signals第66-90页
    4.1. Introduction第66页
    4.2. Problem Description第66-67页
    4.3. Proposed methodology第67-76页
        4.3.1. Training and Enhancement using SNMF第67-71页
        4.3.2. Detection and suppression of typed keystrokes第71-76页
            4.3.2.1 Suppression of Keystrokes by Correlation第72-74页
            4.3.2.2 Detection and replacement by T-F thresholding technique第74-76页
    4.4. Results and Analysis第76-87页
        4.4.1. Signal to Noise Ratio第77-78页
        4.4.2. Perceptual Evaluation of Speech Quality (PESQ)第78页
        4.4.3. Experimental Analysis第78-87页
    4.5. Summary第87-90页
Chapter 5 Conclusion and Future Plan第90-92页
    5.1. Conclusion第90-91页
    5.2. Future Work第91-92页
Chapter 6 References第92-96页
Acknowledgement第96-98页
List of publications第98页

论文共98页,点击 下载论文
上一篇:RNN-BLSTM声学模型的说话人自适应方法研究
下一篇:多人对话场景下的说话人分割聚类研究