Souguil Ann

An EM-based approach for parameter enhancement with an application to speech signals

Abstract This paper considers the estimation of signal parameters and their enhancement using an approach based on the estimation-maximation (EM) algorithm, when only noisy observation data are available. The algorithm is derived with an application to speech signals. The distribution of the excitation source for the speech signal is assumed as a mixture of two Gaussian probability distribution functions with differing variances. This mixture assumption is experimentally valid in enhancing noise-corrupted speech. We recursively estimate the signal parameters and analyze the characteristics of its excitation source in a sequential manner. In the maximum likelihood estimation scheme we utilize the EM algorithm, and employ a detection and an estimation step for the parameters. For their enhancement we use a Kalman filter for the parameters obtained from the estimation procedure. Simulation results using synthetic and real speech data confirm the improved performance of our algorithm in noisy situations, with an increase of about 3 dB in terms of output SNR compared to conventional Gaussian assumption. The proposed algorithm also may be noteworthy in that it needs no voiced/unvoiced decision logic, thanks to the use of the residual approach in the speech signal model.

A fast pitch searching algorithm using correlation characteristics in CELP vocoder

The major drawback in the code excited linear prediction (CELP) type vocoders is their large computational requirements. In the present paper a simple method is proposed to reduce the pitch searching time in the pitch filter almost without degradation of quality. Based upon the observational regularity of the correlation function of speech, the searching range can be restricted to the positive side in pitch search. This is done by skipping the negative side with the width which is estimated from the previous positive envelope. In addition to that, the maximum number of available lags can be limited by the threshold, LT, which is set on 58 empirically. So, only the limited numbers of lags are considered in pitch search, which is less than a half of that of the full search method. By using the proposed method in pitch search, its required computations are greatly reduced. Experimental results show 51% time reduction almost without lowering the speech quality in segmental SNR measures.<<ETX>>

Performance analysis of the dual sign algorithm for additive contaminated-Gaussian noise

In the previous analysis of the dual sign algorithm (DSA), Gaussian signals were assumed. When the desired response contains additive impulsive interference, however, the analysis seems to be inadequate. In this article, a performance analysis of the DSA is considered when the signals are zero-mean stationary Gaussian, and the additive noise of the desired response is zero-mean stationary contaminated-Gaussian (CG). Through computer simulations, our analysis is validated. It is also shown that the DSA is less vulnerable to impulsive interference than the least-mean square (LMS) algorithm.<<ETX>>

Robust estimation of AR parameters and its application for speech enhancement

There are two major problems in estimating vocal tract characteristics by conventional linear prediction: estimation accuracy being subject to the characteristics of the excitation source, and the output quality and the estimation accuracy deteriorating with additive background noise. The authors solved these problems as follows: first, estimate the parameters of a robust AR model where the driving noise is a mixture of a Gaussian and an outlier process; then, propose an iterative procedure that involves parameter estimation for uncorrupted speech and data cleaning based on the robust Kalman filter; lastly, the above results are used to enhance speech corrupted by white noise. The results are more efficient and less biased for uncorrupted speech, and superior at low SNR for noisy speech.<<ETX>>

A classified split vector quantization of LFS parameters

Abstract Split vector quantization (SVQ) of LSF parameter suffers from huge complexity and storage requirements. Increasing the number of subvectors M results in a considerable decrease in both complexity and storage, but at the expense of rapid degradation in performance as M increases. To alleviate this problem, we propose classified SVQ (CSVQ), which uses class-dependent splitting and bit allocation schemes combined with a classified VQ structure. For practical applications, we designed two CSVQ structures. Experimental results have shown that both of the CSVQ schemes achieve nearly transparent quantization at 28 bits/frame while requiring much less complexity than the conventional SVQ.

Adaptive filtering for speech enhancement in colored noise

We consider an adaptive filtering algorithm for speech parameter estimation and enhancement when the observation noise is colored with no a priori information. The resulting algorithm consists of adaptive filtering procedures that recursively estimate and enhance the parameters of speech plus noise model, and results in /spl sim/3 dB improvement over Gaussian assumption on the excitation source.

Robust time-varying parametric modelling of voiced speech

Abstract Although time-varying parametric models have been proposed in the speech literature, few, if any, have had success when applied to speech. Since most employ the LS technique for parameter extraction, these do not perform very well in the presence of large errors (called outliers), which is a distinct possibility when dealing with voiced speech of short-pitch period in particular. In this paper, a robust time-varying LP algorithm is proposed, based on Hubers robust statistics. Testing on both synthetic and natural speech demonstrates that the robust time-varying LP algorithm can eliminate the effect of the glottal excitation on the parameter estimates.

A nonuniform sampling method of speech signal and its application to speech coding

Abstract A nonuniform sampling method for speech signal which rejects perceptually redundant samples by sampling at the maxima and minima of waveform is proposed. Data reduction is improved by silence processing implemented at sampling stage without transmitting any side information. As an application, average 13.4 kbit/s waveform coding scheme is proposed.

A robust algorithm for adaptive FIR filtering and its performance analysis with additive contaminated-Gaussian noise

Abstruct- We introduce a steepest descent linear adaptive algorithm, the proportion-sign algorithm (PSA), lo make the least mean square (LMS) algorithm robust to impulsive interference occurring in the desired response. Its performance analysis is presented when the signals are from zero-mean jlointly stationary Gaussian processes and the additive noise to the (desired response is from a zero-mean stationary contaminated-Gaussian (CG) process which is usually used to represent impulsive interference. Since a special case of the PSA becomes the LMS algorithm, the analysis of the LMS is also obtained as a by-product. By adding a minimal amount of computational complexity, thie PSA improves to some degree the convergence speed over the LMS algorithm without overly degrading the steady-state error performance for Gaussian noise. In addition, since the first derivative of its cost function with respect to estimation error is bounded, it has the properties of robustness to impulsive interference occurring in the desired response while the LMS algorithm is vulnerable to it. Computer simulations are used to demonstrate the validity of our analysis and the robustness of the PSA compared with the LMS algorithm.

Directed graph representation for root-signal set of median filters

Median filtering is a simple digital technique for smoothing signals. A root signal is defined as an invariant signal to the median filtering. We describe directed graph representation for the root-signal set of median filters. The directed graph representation allows us to obtain a set of roots and the number of roots in a straightforward manner.