The performance of adaptive acoustic echo cancelers (AEC) is sensitive to the nonstationarity and correlation of speech signals. In this paper, we explore a new approach based on an adaptive AEC driven by data hidden in speech, to enhance the AEC robustness. We propose a two-stage AEC, where the first stage is a classical NLMS-based AEC driven by the far-end speech. In the signal, we embed-in an extended conception of data hiding-an imperceptible white and stationary signal, i.e., a watermark. The goal of the second stage AEC is to identify the misalignment of the first stage. It is driven by the watermark solely and takes advantage of its appropriate properties (stationary and white) to improve the robustness of the two-stage AEC to the nonstationarity and correlation of speech, and thus reduce the overall system misadjustment. We test two kinds of implementations: in the first implementation, referred to as adaptive watermark driven AEC (A-WdAEC), the watermark is a white stationary Gaussian noise. Driven by this signal, the second stage converges faster than the classical AEC and provides better performance in steady state. In the second implementation, referred to as maximum length sequences WdAEC (MLS-WdAEC), the watermark is built from MLS. Thus, the second stage performs a block identification of the first stage misalignment, given by the circular correlation watermark/preprocessed version of the first stage residual echo. The advantage of this implementation lies in its robustness against noise and undermodeling. Simulation results show the relevance of the “WdAEC” approach, compared to the classical “error-driven AEC.”

Many audio processing algorithms have optimal performance for specific signal statistical distributions that may not be fulfilled for all signals. When the original signal is available, we propose to add an inaudible noise so that the distribution of the signal-plus-noise mixture is as close as possible to a given target distribution. The proposed generic algorithm (independent from the application) adds iteratively a low-power white noise to a flat-spectrum version of the signal, until the target distribution or the noise audibility is reached. The latter is assessed through a frequency masking model. Two implementations of this sound reshaping are described, according to the level of the targeted transformation and to the foreseen application: histogram global reshaping (HGR) to change the global shape of the histogram and histogram local reshaping (HLR) to locally “chisel” the histogram, but keeping the global shape unchanged. These two variants are illustrated by two applications, where the inaudibility of the noise generated by the algorithm is required: “sparsification” for source separation and low-pass filtering of the histogram for application of the quantization theorem, respectively. In both cases, the target histogram is reached or almost reached, and the transformation is inaudible. The experiments show that the source separation performs better with HGR and that the HLR allows a better application of the quantization theorem.

In this paper, we present a temporal entropy-based indicator that reflects the texturedness level of a given audio signal. Inspired from an image homogeneity evaluation via a multidirectional cumulative entropy computation, we similarly propose an audio signal homogeneity analysis through a direct and reverse progressive auditory information content tracking. A [0 - 5] texturedness indicator is then constructed using auditory-inspired parameters, and is inherently associated with short listening time. Using this indicator, a new audio signals classification is proposed where speech signals are assigned low texturedness scores, and academic noise signals are assigned the higher range of texturedness scores. Classically known audio textures are assigned, in this scale, various intermediate to high texturedness scores.

Nonlinear audio system identification generally relies on Gaussianity, whiteness and stationarity hypothesis on the input signal, although audio signals are non-Gaussian, highly correlated and non-stationary. However, since the physical behavior of nonlinear audio systems is input-dependent, they should be identified using natural audio signals (speech or music) as input, instead of artificial signals (sweeps or noise) as usually done. We propose an identification scheme that conditions audio signals to fit the desired properties for an efficient identification. The identification system consists in (1) a Gaussianization step that makes the signal near-Gaussian under a perceptual constraint; (2) a predictor filterbank that whitens the signal; (3) an orthonormalization step that enhances the statistical properties of the input vector of the last step, under a Gaussianity hypothesis; (4) an adaptive nonlinear model. The proposed scheme enhances the convergence rate of the identification and reduces the steady state identification error, compared to other schemes, for example the classical adaptive nonlinear identification.

We consider a geometrical approach for solving the Nonnegative Blind Source Separation (N-BSS) problem in the case of noiseless linear instantaneous mixture model. When the sources are nonnegative, the scatter plot of the mixed data is contained in the simplicial cone generated by the mixing matrix. The proposed method, called Simplicial Cone Shrinking Algorithm for Unmixing Nonnegative Sources (SCSA-UNS), estimates the mixing matrix and the sources by finding the Minimum Volume (MV) simplicial cone containing all the mixed data. Simulations on synthetic data shows the efficiency of the proposed method.

Digital audio watermarking is adopted for many multimedia applications such as copyright management, music indexing, broadcast monitoring, etc. The watermarking techniques must ensure an imperceptible watermark and a reliable transmission with a maximized robustness to a various range of perturbations depending on the application purposes. This paper deals with an improved error correction strategy for digital audio watermarking. It uses a turbo coding algorithm, coupled with a Wiener filtering, that takes into account the equivalent corrupting noise of the watermarking channel and works with reliable soft metrics values. Experimental tests on real audio signals demonstrate improved bit error rates (BER) performances comparing to the reference generic system which uses a Wiener filter at the detection side.

Independent Component Analysis (ICA) is a well-known technique for solving blind source separation (BSS) problem. However “classical” ICA algorithms seem not suited for non-negative sources. This paper proposes a gradient descent approach for solving the Non-Negative Independent Component Analysis problem (NNICA). NNICA original separation criterion contains the discontinuous sign function whose minimization may lead to ill convergence (local minima) especially for sparse sources. Replacing the discontinuous function by a continuous one tanh, we propose a more accurate regularized Gradient algorithm called “Exact” Regularized Gradient (ERG) for NNICA. Experiments on synthetic data with different sparsity degrees illustrate the efficiency of the proposed method and a comparison shows that the proposed ERG outperforms existing methods.

In this paper, we propose the concept of ldquodoping watermarkingrdquo, whose principle is to add an imperceptible noise to an host signal in order to improve its properties. Especially, our aim is to reduce the spectral support of the probability density function (PDF) of an audio signal in order to match the conditions of the quantization theorem. In this context, we develop a specific audio watermarking algorithm and test its performance on real audio signals. This watermark allows to recover the PDF of a digital signal from a sub-quantized version of the signal, with very low error.

This paper evaluates a K-Markov random field model for retrieving information about backscatter characteristics, especially regularity spacing scatterers in simulated ultrasound image. The model combines a statistical K-distribution that describes the envelope of backscattered echo and spatial interaction given by Markov random field (MRF). Parameters estimated by the conditional least squares (CLS) estimation method on simulated radio-frequency (RF) envelope image show that the interaction parameters measure the degree of the randomness of the scatterers

The paper presents a new detection method for a spread spectrum and perceptual watermarking system, viewed as a hidden data transmission system, where the generic detection operation is achieved by a Wiener deconvolution filter. We point out the insufficiencies of the generic reception scheme concerning truncation errors and an ill-conditioned deconvolution operation, and we propose a cascade realization of the reception filter, which significantly improves the detection performances (to a multiplicative factor of 20) at higher bit rates, even in the presence of MPEG compression.

The paper presents an analysis of the perceptual impact of watermarking on the stationarity of audio signals. Indeed, the embedded watermark is piecewise stationary, so it modifies the stationarity of the original audio signal. This study is based on stationarity indices, which represent a measure of the variations in the spectral characteristics of signals, using time frequency representations. Simulation results with two kinds of signals, test signals and audio signals (speech and music) are presented. Comparison stationarity indices between watermarked and original audio signals show a significant stationarity enhancement of the watermarked signal, especially for transient attacks.

This paper presents a new structure of an Acoustic Echo Canceller (AEC) designed to operate in the Mobile Switching Center (MSC) of a GSM network. The purpose of such system is to cancel the echo for all the subscribers. Contrarily to the conventional AEC, the proposed combined AEC/CELP Predictor is able to take into account the non linearities introduced by the GSM speech coders/decoders. A short term predictor is used to model the behavior of the codecs. This new combined system presents higher performance compared to the conventional AEC.

In the hands-free communications, the identification of long impulse response in acoustic echo cancellation requires very important load calculations. One way to reduce the complexity of the classical normalized least mean square (NLMS) adaptive algorithm, is to use the Mmax NLMS algorithm (Aboulnasr and Mayyas 1999). It is shown that this algorithm is a very promising one, that maintains closest performance to the full update NMLS filter in spite of the updating of a small number of coefficients. However, due to its complexity, the mean square analysis uses an unrealistic hypothesis. It was then not possible to consider the practical context such as high input correlation or high step size. In this paper, we present an exact performances analysis, inspired from Besbes et al. (1999), when the input signal remains in a finite alphabet set. With this realistic hypothesis, dedicated to the digital context, we can describe accurately the Mmax NLMS' behavior without any unrealistic assumption. In particular, we evaluate the exact value of critical and optimal step size and we provide the exact mean square error (MSE) for all step sizes and input correlations. The influence of high order statistics can be enhanced.

In this paper, a new class of blind algorithms designed for decision feedback equalization of time varying channels, is proposed. We consider Markovian time variations of the impulse response of the channel as in radio mobile communications. The main idea is to modify classical blind algorithms (decision-directed, constant modulus algorithm,...) in order to give them self-adaptive knowledge of the channel non-stationarity. Simulations show that the proposed algorithms non-stationary DD and non-stationary CMA present better tracking capacity than the classical ones. Hence, they are able to improve the bit error rate especially for severe propagation conditions.

This paper deals with the analysis of adaptive Volterra filters in the finite alphabet inputs case. A tailored approach for this context is presented. Complete and rigorous mean square analysis are provided without any constraining independence assumption. An exact transient and steady state performances such as critical step size, rate of transient decrease, optimal step size and excess mean square error, are deduced.

We propose a new adaptive algorithm designed to track system presented by a filter that has Markovian time evolution. As the non-stationary LMS (NSLMS) algorithm, the non-stationary RLS (NSRLS) algorithm performs better than the LMS and is able to identify the unknown order and parameters of the Markov model. However in the case of the NSRLS algorithm, the convergence speed of the Markovian parameter is very high compared to that of the NSLMS algorithm. Moreover, the NSRLS algorithm has a better tracking capacity than the NSLMS, especially when the filter poles that characterize time variations of the channel are close to the unit circle.

The affine projection algorithm (APA) is a very promising algorithm that has good convergence properties when the input signal is correlated. In particular, it's used to perform communications systems: echo cancellation, equalization. However, due to its complexity, there is no available transient and steady state analysis. We present an exact analysis approach tailored for digital transmission context. In such context, the input signal remains in a finite alphabet set. With a discrete Markov chain model of the inputs, we can describe accurately the APA's behavior without any unrealistic assumption. In particular we calculate the exact value of critical and optimum step size. Moreover, we provide the exact mean square deviation for all step size and input correlation. The influence of high order statistics can be enhanced.

This paper presents a tracking analysis of the LMS algorithm used in order to identify system variations modeled by a random walk. We prove that the steady state properties are strongly related to the input characteristics. The input correlation degrades the performances. Consequently, best performances are obtained for white input. We justify the coupled adaptive predictive structures with system identification in order to improve classical scheme steady state performances.

High speed data transmission over bandlimited and/or dispersive channels require equalization to remove the intersymbol interference (ISI) between the successive transmitted data. This paper presents the cancellation of intersymbol interferences (ISI) as a criterion for the blind adaptation of decision feedback equalizers (DFE). We show that this criterion is an alternative to the decision directed (DD) algorithm. This paper also proposes to replace the hard limiter in the decision device of the DFE by a soft decision implemented by a hyperbolic tangent function in order to escape from local minima during the initialization of the blind algorithms. As the theoretical investigation of these criteria and algorithms is difficult, we analyse some simple examples to illustrate the interest of the proposed solutions.

The unfavorable effects of narrowband inputs on the decoder adjustment when the LMS algorithm is used are analyzed. An explanation is given for the behavior of the CCITT algorithm. Suboptimality of prediction is granted to achieve adjustment, resulting in a satisfactory tradeoff between reduction rate and adjustment. The link between adjustment and uniform stability is enhanced.>

An analysis is presented of the causes of the accumulation of quantizing noise found in the transient state for successive CCITT adaptive differential pulse-code-modulation (ADPCM) transcoders connected synchronously. By decoupling the predictor and quantizer effect it is proved that, owing to a self-stabilization phenomenon, narrowband inputs cause local instabilities in the predictor of the jointly adaptive autoregressive moving-average-prediction/quantization used in the ADPCM 32-kb/s algorithm. Despite the assured global stability, these local instabilities are not synchronized at the encoder and its preceding decoder, and a mistracking occurs which creates quantizing noise accumulation. The tracking is then shown to be very sensitive to predictor/quantizer interaction. The discontinuities introduced in the standardized adaptive quantizer extend the mistracking problem to wideband inputs. A smoothed quantizer with reduced inauspicious interaction is proposed to remedy the problem.>

It is shown that the stochastic error gradient cannot be computed exactly for adaptive recursive filters because of parameter time variations. The usual pseudogradient, (G), which recursively filters the increment, is based on the assumption of slow variations, which is false when the stability boundary is crossed. In the example of adaptive prediction with narrowband inputs, it is shown that (G) may eventually yield an unbounded output. The recursive LMS suppresses the increment filtering and ensures a bounded output, due to the self-stabilization phenomenon, without external stabilization procedures although the adaptive poles may momentarily go outside the unit circle.>