Single-Channel Speech Enhancement

—Noise reduction of speech signals plays an important role in telecommunication systems. Various types of speech additive noise can be introduced such as babble, crowd, large city, and highway which are the main factor of degradation in perceived speech quality. There are some cases on the receiver side of telecommunication systems, where the direct value of interfering noise is not available and there is just access to noisy speech. In these cases the noise cannot be cancelled totally but it may be possible to reduce the noise in a sensible way by utilizing the statistics of the noise and speech signal. In this paper the proposed method for noise reduction is Bayesian recursive state-space Kalman filter, which is a method for estimation of a speech signal from its noisy version. It utilizes the prior probability distributions of the signal and noise processes, which are assumed to be zero-mean Gaussian processes. The function of Kalman filter is assessed for different types of noise such as babble, crowd, large city, and highway. The noise cancellation is implemented for each of aforementioned noises which their powers vary in a range of values. This method of noise reduction yields better speech perceived quality and efficient results compared to Wien-er filter.

We introduce a non-negative matrix factorization technique which learns speech features with temporal extent in the presence of non-stationary noise. Our proposed technique, namely Sparse convolutive robust non-negative matrix factorization, is robust in the presence of noise due to our explicit treatment of noise as an interfering source in the factorization. We derive multiplicative update rules using the alpha divergence objective. We show that our proposed method yields superior performance to sparse convolutive non-negative matrix factorization in a feature learning task on noisy data and comparable results to dedicated speech enhancement techniques.

— The noise exists in almost all environments such as cellular mobile telephone systems. Various types of noise can be introduced such as speech additive noise which is the main factor of degradation in perceived speech quality. At some applications for example at the receiver of a telecommunication system, the direct value of interfering noise is not available and there is just access to noisy speech. In these cases the noise cannot be cancelled totally but it may be possible to reduce the noise in a sensible way by utilizing the statistics of the noise and speech signal. In this paper the proposed method for noise reduction is Bayesian recursive state-space Kalman filter, which is a method for estimation of a speech signal from its noisy version. It utilizes the prior probability distributions of the signal and noise processes, which are assumed to be zero-mean Gaussian processes, to implement the noise reduction. The function of Kalman filter is assessed for SNR values of-5 dB and 5 dB respectively. Evaluation results indicate that this method of noise reduction yields better speech perceived quality and efficient results.

We introduce a non-negative matrix factorization technique which learns speech features with temporal extent in the presence of non-stationary noise. Our proposed technique, namely Sparse convolutive robust non-negative matrix factorization, is robust in the presence of noise due to our explicit treatment of noise as an interfering source in the factorization. We derive multiplicative update rules using the alpha divergence objective. We show that our proposed method yields superior performance to sparse convolutive non-negative matrix factorization in a feature learning task on noisy data and comparable results to dedicated speech enhancement techniques.

@inproceedings {deFrein09Learning,
  title            = {{Learning Speech Features in the Presence of Noise: Sparse Convolutive Robust Non-negative Matrix Factorisation}},
  year            = "2009",
  author          = "\href{http://robustandscalable.wordpress.com}{Ruair\'{i} de Fr\'{e}in} and Scott T. Rickard",
  booktitle        = "Digital Signal Processing, 2009 16th International Conference on",
  pages            = "1--6",
  organization    = "IEEE",
  doi              = "10.1109/ICDSP.2009.5201068",
  urllink          = "http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=5201068&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D5201068",
  keywords        = "matrix algebra;speech enhancement;alpha divergence objective;nonstationary noise;sparse convolutive robust non-negative matrix factorization;speech enhancement techniques;speech features;Additive noise;Background noise;Matrix decomposition;Noise reduction;Noise robustness;Sparse matrices;Spectrogram;Speech analysis;Speech enhancement;Working environment noise;Spectral factorization;Speech enhancement",
  month            = "Jul."}