Statistics Of Speech Signals Research Articles

Single Channel Speech Enhancement: Using Wiener Filtering with Recursive Noise Estimation

This paper discusses the problem of single channel speech enhancement in stationary environments, and proposes Wiener filtering with the recursive noise estimation algorithm. The Wiener filter is a linear estimator and minimizes the mean-squared error between the original and enhanced speech. The algorithm is implemented in the frequency domain and depends on the filter transfer function from sample to sample based on the speech signal statistics; the local mean and the local variance. For the noise estimation, the recursive noise estimation approach is used. In this approach, the noise estimation is done by past and present spectral power values, using a smoothing parameter. The value of smoothing parameter is selected in between [0 1]. For the performance evaluation of the proposed speech enhancement algorithm objective evaluations with informal listening tests are conducted for the speech sentences, pronounced by male and female speakers from the NOIZEUS corpus, degraded by White as well as Pink noise types at different SNR levels. For objective measures, signal to noise ratio, segmental signal to noise ratio, and the perceptual evaluation of speech quality are used. The measures prove that the speech enhanced by proposed algorithm is more pleasant to the human ear for both noise conditions in comparison to the conventional speech enhancement method.

Speech enhancement with an adaptive Wiener filter

This paper proposes an adaptive Wiener filtering method for speech enhancement. This method depends on the adaptation of the filter transfer function from sample to sample based on the speech signal statistics; the local mean and the local variance. It is implemented in the time domain rather than in the frequency domain to accommodate for the time-varying nature of the speech signals. The proposed method is compared to the traditional frequency-domain Wiener filtering, spectral subtraction and wavelet denoising methods using different speech quality metrics. The simulation results reveal the superiority of the proposed Wiener filtering method in the case of Additive White Gaussian Noise (AWGN) as well as colored noise.

Enhancing speech using an Adaptive Wiener Filter based algorithm

Speech can be expressed as a mechanism of expressing thoughts and ideas using vocal sounds. In humans, speech signals are generated at vocal cords, travelled through the vocal tract, and finally produced & transmitted through speaker’s mouth. These speech signals then usually travels through air or other mediums to the listener’s ear, where it acts as pressure waves. The bandwidth of speech signal is around 4 KHz. The noise produced by various ambient sources such as vehicles normally lies in this frequency range. Therefore, speech signals get easily distorted by the ambient noise. These distorted or degraded speech signals are called noisy speech signals. This paper focuses on speech processing (in particularly speech enhancement) of the noisy speech signals. This is very important as speech is the most commonly used way of communication and interaction between humans; however, it is very complex to understand. Therefore, this paper proposes an adaptive algorithm based on Wiener filter for speech enhancement. The proposed adaptive Wiener filter depends on the adaptation of the filter transfer function from sample to sample based on the speech signal statistics (mean and variance). The adaptive Wiener filter is implemented in time domain rather than in frequency domain. This is done to accommodate the random speech signal. The proposed method is compared to the traditional Wiener filter and the spectral subtraction methods.

A Lattice Transversal Joint Adaptive Filter with Fixed Reflection Coefficients

We present a lattice transversal joint (LTJ) adaptive filter with fixed reflection coefficients to achieve fast convergence with low complexity. The reflection coefficients of the filter are given by the statistics of speech signals, and the proposed order of the lattice predictor is one. Experimental results confirm that as compared to the adaptive transversal filter, the proposed adaptive filter achieves fast convergence with a negligible increase in complexity. The proposed adaptive filter converges around six times faster than the adaptive transversal filter in case of the band-limited voiced signal from the ITU-T G.168 standard.

Speech probability distribution

It is demonstrated that the distribution of speech samples is well described by Laplacian distribution (LD). The widely known speech distributions, i.e., LD, Gaussian distribution (GD), generalized GD, and gamma distribution, are tested as four hypotheses, and it is proved that speech samples during voice activity intervals are Laplacian random variables. A decorrelation transformation is then applied to speech samples to approximate their multivariate distribution. To do this, speech is decomposed using an adaptive Karhunen-Loeve transform or a discrete cosine transform. Then, the distributions of speech components in decorrelated domains are investigated. Experimental evaluations prove that the statistics of speech signals are like a multivariate LD. All marginal distributions of speech are accurately described by LD in decorrelated domains. While the energies of speech components are time-varying, their distribution shape remains Laplacian.

Letter: Optimal echo path modelling using chaotic modulated speech

AbstractA novel modulating scheme is presented that allows optimal echo path modelling using, real time least mean square (LMS) system identification techniques. The systems of interest are usually based on a class of digital adaptive filters (DAFs). The modelling convergence rate derived from the optimal Wiener weights defines the performance criterion. A novel chaotic based modulation regime utilising the logistic mapping is exploited to whiten the speech power spectral density (PSD) whilst preserving the signal bandwidth requirements. The whitened signal optimises the convergence rate and allows adaptation times to be calculated without a‐priori knowledge of the speech signal's second order statistics. Software simulations are reported for an ensemble of conversational speech realisations and it is shown that the described modulation scheme significantly improves the convergence rate and guarantees optimal modelling independent of the input speech signal statistics. Furthermore, the noise whitened input signal circumvents control supervisory problems usually associated with uncoded speech inputs.

Authors' replies to the discussion on “A p.c.m. logarithmic encoder for a multi-channel t.d.m. system” and “A survey of telephone speech-signal statistics and their significance in the choice of a p.c.m. companding law”

Authors' replies to the discussion on “A p.c.m. logarithmic encoder for a multi-channel t.d.m. system” and “A survey of telephone speech-signal statistics and their significance in the choice of a p.c.m. companding law”

Discussion on “A p.c.m. logarithmic encoder for a multi-channel t.d.m. system”, by J.C.H. Davis, “A reference distortion system using modulated noise”, by H.B. Law, and “A survey of telephone speech-signal statistics and their significance in the choice of a p.c.m. companding law”, by R.F. Purton, before the Electroncs and Communications Section, 15th May, 1962

Discussion on “A p.c.m. logarithmic encoder for a multi-channel t.d.m. system”, by J.C.H. Davis, “A reference distortion system using modulated noise”, by H.B. Law, and “A survey of telephone speech-signal statistics and their significance in the choice of a p.c.m. companding law”, by R.F. Purton, before the Electroncs and Communications Section, 15th May, 1962

Survey of telephone speech-signal statistics

Survey of telephone speech-signal statistics

A survey of telephone speech-signal statistics and their significance in the choice of a p.c.m. companding law

Published information concerning the mean speech volume and likely range of volumes in a telephone circuit is summarized, and curves are reproduced for the variation in instantaneous amplitude of an individual talker. It is concluded that, to be able to handle both types of variation, a p.c.m. system should have its smallest quantization step at least 62dB below the peak limiting level. This range can be handled by a 7-digit binary code using non-linear encoding (companding). Two main types of non-linear characteristic, namely logarithmic and hyperbolic, are compared, first on the basis of signal/quantizing-noise ratio and secondly in terms of theoretical information content. It is shown that, provided that the non-linearity parameter is chosen to give an adequate dynamic range, there is little to choose between the two characteristics. The possibility of using automatic volume regulators to reduce the dynamic range is also considered, but it is concluded that they give little advantage.