栏目导航

 通知公告 
 图片新闻 
 学院新闻 
 科学方舟 
 学院荣誉 
 学术动态 
  • “双一流”建设

    更多>>
  • 当前位置: 首页>>学术动态>>正文
    以色列理工学院的Israel Cohen 教授报告:Title: New Statistical Model for the Enhancement of Noisy Speech
     

    以色列理工学院的Israel Cohen 教授报告

    Talk::Title: New Statistical Model for the Enhancement of Noisy Speech

     
    主持:陈景东教授
    时间:2013年11月5日9:00
    地点:航海学院201会议室
      Abstract:Modeling speech in the short-time Fourier transform (STFT) domain is a
    fundamental problem in designing speech enhancement systems. A major
    drawback of existing Gaussian and supergaussian speech models is that variances
    of the spectral coefficients are estimated from the noisy observed signal by the
    decision-directed approach. The decision-directed approach is not supported by a
    statistical model, and hence cannot be adapted to the components of the speech
    signal. In fact, the parameters of the decision-directed estimator have to be
    determined in advance by simulations and subjective listening tests for each
    particular setup of time-frequency transformation and speech enhancement
    algorithm. Statistical models based on hidden Markov models (HMMs) try to
    circumvent the assumption of specific distributions. However, the HMM-based
    speech enhancement relies on the type of training data. It works best with the
    trained type of noise, but often worse with other type of noise. Furthermore,
    improved performance generally entails more complex models and higher
    computational requirements.

    In this talk, we introduce a novel modeling approach, which is based on
    generalized autoregressive conditional heteroscedasticity (GARCH). GARCH
    models are widely-used in various financial applications such as risk
    management, option pricing and the term structure of interest rates. They
    explicitly parameterize the time-varying volatility in terms of past conditional
    variances and past squared innovations, while taking into account excess kurtosis
    and volatility clustering, two important characteristics of financial time-series. We
    show that speech signals in the STFT domain are also characterized by heavy
    tailed distributions and volatility clustering. We describe speech enhancement
    algorithms, which are based on GARCH modeling, and demonstrate their
    improved performance compared to using conventional supergaussian modeling
    and the decision-directed method.

    Speaker:  Israel Cohen 

    Israel Cohen  is a professor of electrical engineering at the Technion–Israel 
    Institute of Technology, Heifa, Israel. He received the B.Sc. (summa cum 
    laude), M.Sc., and Ph.D. degrees in electrical engineering from the Technion–
    Israel Institute of Technology in 1990, 1993, and 1998, respectively.
    From 1990 to 1998, he was a research scientist with
    RAFAEL Research Laboratories, Haifa, Israel Ministry of Defense.
    From 1998 to 2001, he was a postdoctoral research associate with
    the Computer Science Department, Yale University, New Haven,
    Connecticut. In 2001, he joined the Electrical Engineering
    Department, Technion. His research interests are statistical signal
    processing, analysis and modeling of acoustic signals,
    speech enhancement, noise estimation, microphone arrays,
    source localization, blind source separation, system identification,
    and adaptive filtering. He is a coeditor of the Multichannel
    Speech Processing section of the Springer Handbook of Speech
    Processing (Springer, 2008), a coauthor of Noise Reduction in
    Speech Processing (Springer, 2009), a coeditor of Speech Processing
    in Modern Communication: Challenges and Perspectives
    (Springer, 2010), and a general cochair of the 2010
    International Workshop on Acoustic Echo and Noise Control.
    He received the Alexander Goldberg Prize for Excellence in
    Research and the Muriel and David Jacknow Award for Excellence
    in Teaching. He is a member of the IEEE Audio and
    Acoustic Signal Processing Technical Committee and the IEEE
    Speech and Language Processing Technical Committee. He was
    an associate editor of IEEE Transactions on Audio, Speech, and
    Language Processing and IEEE Signal Processing Letters. He
    was a guest editor of a special issue of the European Association
    for Signal Processing’s (EURASIP’s) Journal on Advances in
    Signal Processing on “Advances in Multimicrophone Speech
    Processing” and a special issue of Elsevier’s Speech Communication
    journal on “Speech Enhancement.”