Advanced Search
Volume 30 Issue 2
Jan.  2011
Turn off MathJax
Article Contents
Pan Xin-yu, Zhao He-ming, Chen Xue-qin, Xu Min . Endpoint Detection of Whispers Based on the Fitting Characteristic of EMD[J]. Journal of Electronics & Information Technology, 2008, 30(2): 362-366. doi: 10.3724/SP.J.1146.2006.01021
Citation: Pan Xin-yu, Zhao He-ming, Chen Xue-qin, Xu Min . Endpoint Detection of Whispers Based on the Fitting Characteristic of EMD[J]. Journal of Electronics & Information Technology, 2008, 30(2): 362-366. doi: 10.3724/SP.J.1146.2006.01021

Endpoint Detection of Whispers Based on the Fitting Characteristic of EMD

doi: 10.3724/SP.J.1146.2006.01021
  • Received Date: 2006-07-10
  • Rev Recd Date: 2007-01-12
  • Publish Date: 2008-02-19
  • Whispered speech is the especial form of peoples pronunciation. There is lower Signal-to-Noise Ratio (SNR) in whispers and unobvious pitch waveform compared with the normal speech, so it is more difficult to process the whispered speech. The endpoint detection of whispers is the first pivotal step of whispered speech signal processing. This paper uses the Empirical Mode Decomposition (EMD) of Hilbert-Huang Transform (HHT) to solve the problem, and firstly proposes a novel algorithm of endpoint detection of whispered speech based on the fitting characteristic of EMD. Normalize the energy of Intrinsic Mode Function (IMF) obtained by EMD, and use the fitting parameters of the energy as the characteristic and then the endpoint of whispers can be easily divided. The results of experiments show that it is very useful in endpoint detection of whispers, and the accurate rate is 98.25% in 1200 samples (SNR=2~10dB).
  • loading
  • 陈四根, 和应民. 一种基于信息熵的语音端点检测方法 [J].应用科技, 2001, 28(3): 13-14.Chen S G and He Y M. A scheme of speech endpointdetection based on information entropy [J]. Applied Scienceand Technology, 2001, 28(3): 13-14.[2]Drouiche K, Gomez P, Alvarez A, Martinez R, Rodellar V,and Nieto V. A spectral distance measure for speechdetection in noise and speech segmentation [C]. Proceedingsof the 11th IEEE Signal Processing Workshop on StatisticalSignal Processing, Singapore, 2001: 500-503.[3]Chen S H, Liao Y F, and Chiang S M, et al.. An RNN-basedpre-classification method for fast continuous mandarinspeech recognition [J].IEEE Trans. on Speech and AudioProcessing.1998, 6(1):86-90[4]Robert W M and Mark A C. Reconstruction of speech fromwhispers [J].Medical Engineering Physics.2002, 24(8):515-520[5]栗学丽, 丁慧, 徐柏龄. 基于熵函数的耳语音声韵分割法 [J].声学学报, 2005, 30(1): 69-75.Li X L, Ding H, and Xu B L. Entropy-based initial/finalsegmentation for Chinese whispered speech [J]. ActaAcoustica, 2005, 30(1): 69-75.[6]Liu Z F, Liao Z P, and Sang E F. Speech enhancement basedon Hilbert-Huang transform [C]. Proceedings of 2005International Conference on Machine Learning andCybernetics, Guangzhou, China, 2005, 8: 4908-4912.[7]杨莉莉, 李燕, 徐柏龄. 汉语耳语音库的建立与听觉实验研究[J]. 南京大学学报(自然科学), 2005, 41(3): 311-317.Yang L L, Li Y, and Xu B L. The establishment of a Chinesewhisper database and perceptual experiment [J]. Journal ofNanjing University (Natural Sciences), 2005, 41(3): 311-317.[8]吴宗济,林茂灿主编. 实验语音学概要[M]. 北京: 高等教育出版社, 1989: 112-152.Wu Z J and Lin M C. Experiment Phonetics [M]. Beijing:Higher Education Press, 1989: 112-152.
  • 加载中

Catalog

    通讯作者: 陈斌, bchen63@163.com
    • 1. 

      沈阳化工大学材料科学与工程学院 沈阳 110142

    1. 本站搜索
    2. 百度学术搜索
    3. 万方数据库搜索
    4. CNKI搜索

    Article Metrics

    Article views (3168) PDF downloads(1067) Cited by()
    Proportional views
    Related

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return