Search

The analysis of (approximately) periodic signals is an importantelement in numerous applications. One generalization of standardperiodic signals often occurring in practice are harmonic chirpsignals where the instantaneous frequency increases/decreases linearlyas a function of time. A statistically efficient estimator forextracting the parameters of the harmonic chirp model in additivewhite Gaussian noise is the maximum likelihood (ML) estimator whichrecently has been demonstrated to be robust to noise and accurate ---even when the model order is unknown. The main drawback of the MLestimator is that only very computationally demanding algorithms forcomputing an estimate are known. In this paper, we give an algorithmfor computing an estimate to the ML estimator for a number ofcandidate model orders with a much lower computational complexity thanpreviously reported in the literature. The lower computationalcomplexity is achieved by exploiting recursive matrix structures,including a block Toeplitz-plus-Hankel structure, the fast Fouriertransform, and using a two-step approach composed of a grid andrefinement step to reduce the number of required functionevaluations. The proposed algorithms are assessed via Monte Carlo andtiming studies. The timing studies show that the proposed algorithm isorders of magnitude faster than a recently proposed algorithm forpractical sizes of the number of harmonics and the length of thesignal.

Modern-day hearing aids are capable of receiving acoustic signals over a wireless link and also from the surroundings through the microphone. If the hearing aid receives input only from the acoustic environment, feedback cancellation proceeds according to the existing methodologies for bias reduction. However, the wirelessly received signal and the acoustic environment input, when emitted from the same source, can be very similar to each other or with a time-delayed version of each other, thereby having a high correlation between them. Both inputs can also be emitted from different sources and, thus, be less correlated with each other. In the aforementioned scenarios, acoustic confusion can occur for the user as the hearing aid receives both signals simultaneously. To improve the output signal quality and to reduce bias in an adaptive feedback cancellation system with a wirelessly received signal as well as an acoustic environment input, we propose a cost function, and the optimization of the feed-forward path and of the shaping filter for the wireless signal. The feed-forward path is designed to be a cascade of the required acoustic enhancement along with an FIR filter. We derive expressions for an optimum shaping filter and for an optimized feed-forward path. Improvement in loudspeaker output signal quality, normalized misalignment and maximum stable gain for each of the above-mentioned scenarios is assessed through numerical simulations.

We consider an adaptive linear prediction based feedback canceller for hearing aids that exploits two (an external and a shaped) noise signals for a bias-less adaptive estimation. In particular, the bias in the estimate of the feedback path is reduced by synthesizing the high-frequency spectrum of the reinforced signal using a shaped noise signal. Moreover, a second shaped (probe) noise signal is used to reduce the closed-loop signal correlation between the acoustic input and the loudspeaker signal at low frequencies. A power-transfer-function analysis of the system is provided, from which the effect of the system parameters and adaptive algorithms [normalized least mean square (NLMS) and recursive least square (RLS)] on the rate of convergence, the steady-state behaviour and the stability of the feedback canceller is explicitly found. The derived expressions are verified through computer simulations. It is found that, as compared to feedback canceller without probe noise, the cost of achieving an unbiased estimate of the feedback path using the feedback canceller with probe noise is a higher steady-state misadjustment for the RLS algorithm, whereas a slower convergence and a higher tracking error for the NLMS algorithm.

This letter proposes a real-time perceptual model predicting the experienced distraction occurring in interfering audio-on-audio situations. The proposed model improves the computational efficiency of a previous distraction model, which cannot provide predictions in real time. The chosen approach was to utilize similar features as the previous model, but to use faster underlying algorithms to calculate these features. The results show that the proposed model has a root mean squared error of 11.9%, compared to the previous model's 11.0%, while only taking 0.04% of the computational time of the previous model. Thus, while providing similar accuracy as the previous model, the proposed model can be run in real time. The proposed distraction model can be used as a tool for evaluating and optimizing sound-zone systems. Furthermore, the real-time capability of the model introduces new possibilities, such as adaptive sound-zone systems.

Modelling signals as being periodic is common in many applications. Such periodic signals can be represented by a weighted sum of sinusoids with frequencies being an integer multiple of the fundamental frequency. Due to its widespread use, numerous methods have been proposed to estimate the fundamental frequency, and the maximum likelihood (ML) estimator is the most accurate estimator in statistical terms. When the noise is assumed to be white and Gaussian, the ML estimator is identical to the non-linear least squares (NLS) estimator. Despite being optimal in a statistical sense, the NLS estimator has a high computational complexity. In this paper, we propose an algorithm for lowering this complexity significantly by showing that the NLS estimator can be computed efficiently by solving two Toeplitz-plus-Hankel systems of equations and by exploiting the recursive-in-order matrix structures of these systems. Specifically, the proposed algorithm reduces the time complexity to the same order as that of the popular harmonic summation method which is an approximate NLS estimator. The performance of the proposed algorithm is assessed via Monte Carlo and timing studies. These show that the proposed algorithm speeds up the evaluation of the NLS estimator by a factor of 50–100 for typical scenarios.

This paper introduces the concept of affective offset, which is the difference between a user's perceived a ffective state and the affective annotation of the content they wish to see. We show how this affective offset can be used within a framework for providing recommendations for TV progr ams. First a user's mood profile is determined using 12-class audio-based emotion classifications. An initial TV content item is then displayed to the user based on the extracted mood profil e. The user has the option to either accept the recommendation, or to critique the item once or several times, by navigating the emotion space to request an alternative match. Thefina lm atch is then compared to the initial match, in terms of the difference in the items' affective parameterization. This offset is then utilized in future recom- mendation sessions. The system was evaluated by eliciting three different moods in 22 separate users and examining the influence of applying affective offset to th eu sers' sessions. Results show that, in the case when affecti ve offset was applied, better user satisfaction was achieved: the average ratings went from 7.80 up to 8.65, with an average decrease in the number of critiquing cycles which went from 29 .53 down to 14.39. Index Terms—Affective offset, circumplex model of affect, critique-based recommenders, emotions, EPG, moods.

Acoustic feedback is a well-known problem in hearing aids, caused by the undesired acoustic coupling between the hearing aid loudspeaker and microphone. Acoustic feedback produces annoying howling sounds and limits the maximum achievable hearing aid amplification. This paper is focused on adaptive feedback cancellation (AFC) where the goal is to adaptively model the acoustic feedback path and estimate the feedback signal, which is then subtracted from the microphone signal. The main problem in identifying the acoustic feedback path model is the correlation between the near-end signal and the loudspeaker signal caused by the closed signal loop, in particular when the near-end signal is spectrally colored as is the case for a speech signal. This paper adopts a prediction-error method (PEM)-based approach to AFC, which is based on the use of decorrelating prediction error filters (PEFs). We propose a number of improved PEF designs that are inspired by harmonic sinusoidal modeling and pitch prediction of speech signals. The resulting PEM-based AFC algorithms are evaluated in terms of the maximum stable gain (MSG), filter misadjustment, and computational complexity. Simulation results for a hearing aid scenario indicate an improvement up to 5–7 dB in MSG and up to 6–8 dB in terms of filter misadjustment.

We consider a source coding problem with a network scenario in mind, and formulate it as a remote vector Gaussian Wyner-Ziv problem under covariance matrix distortions. We define a notion of minimum for two positive-definite matrices based on which we derive an explicit formula for the rate-distortion function (RDF). We then study the special cases and applications of this result. We show that two well-studied source coding problems, i.e. remote vector Gaussian Wyner- Ziv problems with mean-squared error and mutual information constraints are in fact special cases of our results. Finally, we apply our results to a joint source coding and denoising problem. We consider a network with a centralized topology and a given weighted sum-rate constraint, where the received signals at the center are to be fused to maximize the output SNR while enforcing no linear distortion. We show that one can design the distortion matrices at the nodes in order to maximize the output SNR at the fusion center. We thereby bridge between denoising and source coding within this setup.

In many existing time-domain filtering methods for noise reduction in, e.g., speech processing, the filters are causal. Such causal filters can be implemented directly in practice. However, it is possible to improve the performance of such noise reduction filtering methods in terms of both noise suppression and signal distortion by allowing the filters to be non-causal. Non-causal time-domain filters require knowledge of the future, and are therefore not directly implementable. If the observed signal is processed in blocks, however, the non-causal filters are implementable. In this paper, we propose such non-causal time-domain filters for noise reduction in speech applications. We also propose some performance measures that enable us to evaluate the performance of non-causal filters. Moreover, it is shown how some of the filters can be updated recursively. Using the recursive expressions, it is also shown that the output SNRs of the filters always increase as we increase the length of the filter when the desired signal is stationary. From both the theoretical and practical evaluations of the filters, it is clearly shown that the performance of time-domain filtering methods for noise reduction can be improved by introducing non-causality.

Noise reduction (NR) and dynamic range compression (DRC) are basic components in hearing aids, but generally these components are developed and evaluated independently of each other. Hearing aids typically use a serial concatenation of NR and DRC. However, the DRC in such a concatenation negatively affects the performance of the NR stage: the residual noise after NR receives more amplification compared to the speech, resulting in a signal-to-noise-ratio (SNR) degradation. The integration of NR and DRC has not received a lot of attention so far. In this paper, a multi-channel Wiener filter (MWF)-based approach is presented for speech and noise scenarios, where an MWF-based NR algorithm is combined with DRC. The proposed solution is based on modifying the MWF and the DRC to incorporate the conditional speech presence probability in order to avoid residual noise amplification. The goal is then to analyse any undesired interaction effects by means of objective measures. Experimental results indeed confirm that a serial concatenation of NR and DRC degrades the SNR improvement provided by the NR, whereas the combined approach proposed here shows less degradation of the SNR improvement at a low increase in distortion compared to a serial concatenation.

Encouraged by the promising application of compressed sensing in signal compression, we investigate its formulation and application in the context of speech coding based on sparse linear prediction. In particular, a compressed sensing method can be devised to compute a sparse approximation of speech in the residual domain when sparse linear prediction is involved. We compare the method of computing a sparse prediction residual with the optimal technique based on an exhaustive search of the possible nonzero locations and the well known Multi-Pulse Excitation, the first encoding technique to introduce the sparsity concept in speech coding. Experimental results demonstrate the potential of compressed sensing in speech coding techniques, offering high perceptual quality with a very sparse approximated prediction residual.

Udgivelsesdato: NOV We present a new design method for robust low-delay coding of autoregressive sources for transmission across erasure channels. It is a fundamental rethinking of existing concepts. It considers the encoder a mechanism that produces signal measurements from which the decoder estimates the original signal. The method is based on linear predictive coding and Kalman estimation at the decoder. We employ a novel encoder state-space representation with a linear quantization noise model. The encoder is represented by the Kalman measurement at the decoder. The presented method designs the encoder and decoder offline through an iterative algorithm based on closed-form minimization of the trace of the decoder state error covariance. The design method is shown to provide considerable performance gains, when the transmitted quantized prediction errors are subject to loss, in terms of signal-to-noise ratio (SNR) compared to the same coding framework optimized for no loss. The design method applies to stationary auto-regressive sources of any order. We demonstrate the method in a framework based on a generalized differential pulse code modulation encoder. The presented principles can be applied to more complicated coding systems that incorporate predictive coding as well.

Udgivelsesdato: APR

Knowledge on how a number of loudspeakers are positioned relative to a listening position can be used to enhance the listening experience. Usually, these loudspeaker positions are estimated using calibration signals, either audible or psycho-acoustically hidden inside the desired audio signal. In this paper, we propose to use the desired audio signal instead. Specifically, we treat the case of estimating the distance between two loudspeakers playing back a stereo music or speech signal. In this connection, we develop a real-time maximum likelihood estimator and demonstrate that it has a variance in the millimetre range in a real environment for even a modest sampling frequency.

A fundamental issue in connection with subspace methods for noise reduction is that the covariance matrix for the noise is required to have full rank in order for the prewhitening step to be defined. However, there are important cases where this requirement is not fulfilled, e.g., when the noise has narrowband characteristics or in the case of tonal noise. We extend the concept of prewhitening to include the case when the noise covariance matrix is rank deficient, using a weighted pseudoinverse and the quotient singular value decomposition, and we show how to formulate a general rank-reduction algorithm that works also for rank-deficient noise. We also demonstrate how to formulate this algorithm by means of a quotient ULV decomposition, which allows for faster computation and updating. Finally, we apply our algorithm to a problem involving a speech signal contaminated by narrowband noise.

In this paper, we consider the problem of source coding for a wireless acoustic sensor network where each node in the network makes its own noisy measurement of the sound field, and communicates with other nodes in the network by sending and receiving encoded versions of the measurements. To make use of the correlation between the sources available at the nodes, we consider the possibility of combining the measurement and the received messages into one single message at each node instead of forwarding the received messages and separate encoding of the measurement. Moreover, to exploit the correlation between the messages received by a node and the node?s measurement of the source, we propose to use the measurement as side information and thereby form a distributed source coding (DSC) problem. Assuming that the sources are Gaussian, we then derive the rate-distortion function (RDF) for the resulting remote DSC problem under covariance matrix distortion constraints. We further show that for this problem, the Gaussian source is the worst to code. Thus, the Gaussian RDF provides an upper bound to other sources such as audio signals. We then turn our attention to audio signals. We consider an acoustical model based on the room impulse response (RIR) and provide simulation results for the rate-distortion performance in a practical setup where a set of microphones record the sound in a standard listening room. Since our reconstruction scheme and distortion measure are defined over the direct sound source, coding and dereverberation are performed in a joint manner. HighlightsWe treat the problem of source coding for wireless acoustic sensor networks.We consider vector sources to make use of the time correlation in the audio sequences.We use the measurements at the receiving nodes as side information using distributed source coding.We derive local rate-distortion functions to be used for rate allocation for an optimal sum-rate.Our encoding/decoding process is joint with dereverberation.

We propose new compressive parameter estimation algorithms that make use of polar interpolation to improve the estimator precision. Our work extends previous approaches involving polar interpolation for compressive parameter estimation in two aspects: (i) we extend the formulation from real non-negative amplitude parameters to arbitrary complex ones, and (ii) we allow for mismatch between the manifold described by the parameters and its polar approximation. To quantify the improvements afforded by the proposed extensions, we evaluate six algorithms for estimation of parameters in sparse translation-invariant signals, exemplified with the time delay estimation problem. The evaluation is based on three performance metrics: estimator precision, sampling rate and computational complexity. We use compressive sensing with all the algorithms to lower the necessary sampling rate and show that it is still possible to attain good estimation precision and keep the computational complexity low. Our numerical experiments show that the proposed algorithms outperform existing approaches that either leverage polynomial interpolation or are based on a conversion to a frequency-estimation problem followed by a super-resolution algorithm. The algorithms studied here provide various tradeoffs between computational complexity, estimation precision, and necessary sampling rate. The work shows that compressive sensing for the class of sparse translation-invariant signals allows for a decrease in sampling rate and that the use of polar interpolation increases the estimation precision., Comment: 13 pages, 5 figures, to appear in IEEE Transactions on Signal Processing; minor edits and corrections

Speaker de-identification is an interesting and newly investigated task in speech processing. In the current implementations, this task is based on transforming one speaker speech to another speaker in order to hide the speaker identity. In this paper we present a discriminative approach for human speaker selection for speaker de-identiication. We used two modules, a speaker identiication system and a speaker transformation one, to select the most appropriate speaker to transform the source speaker speech from a set of speakers. In order to select the target speaker, we minimize the identi-ication conidence of the transformed speech as the source speaker and maximize the confusion about the transformed speech membership to the rest of the speaker models and the identiication conidence of the re-transformed speech using the source speaker model. These three factors are combined to achieve overall optimization performance in order to select the best target speaker to transform the source.

In the signal processing literature, many methods have been proposed for estimating the number of sinusoidal basis functions from a noisy data set. The most popular method is the asymptotic MAP criterion, which is sometimes also referred to as the BIC. In this paper, we extend and improve this method by considering the problem in a full Bayesian framework instead of the approximate formulation, on which the asymptotic MAP criterion is based. This leads to a new model selection and comparison method, the lp-BIC, whose computational complexity is of the same order as the asymptotic MAP criterion. Through simulations, we demonstrate that the lp-BIC outperforms the asymptotic MAP criterion and other state of the art methods in terms of model selection, de-noising and prediction performance. The simulation code is available online.

We show that the reduced-rank output signal computed via truncated (Q)SVD is identical to that from an array of parallelly connected analysis-synthesis finite impulse response (FIR) filter pairs. The filter coefficients are determined by the (Q)SVD, and the filters provide an explicit description of the reduced-rank noise reduction algorithm in the frequency domain.

Model comparison and selection is an importantproblem in many model-based signal processing applications.Often, very simple information criteria such as the Akaikeinformation criterion or the Bayesian information criterion areused despite their shortcomings. Compared to these methods,Djuric’s asymptotic MAP rule was an improvement, and in thispaper we extend the work by Djuric in several ways. Specifically,we consider the elicitation of proper prior distributions, treatthe case of real- and complex-valued data simultaneously ina Bayesian framework similar to that considered by Djuric,and develop new model selection rules for a regression modelcontaining both linear and non-linear parameters. Moreover,we use this framework to give a new interpretation of thepopular information criteria and relate their performance to thesignal-to-noise ratio of the data. By use of simulations, we alsodemonstrate that our proposed model comparison and selectionrules outperform the traditional information criteria both interms of detecting the true model and in terms of predictingunobserved data. The simulation code is available online.

Double-talk (DT)-robust acoustic echo cancellation (AEC) and acoustic feedback cancellation (AFC) are needed in speech communication systems, e.g., in hands-free communication systems and hearing aids. In this paper, we derive a practical and computationally efficient algorithm based on the frequency-domain adaptive filter prediction error method using row operations (FDAF-PEM-AFROW) for DT-robust AEC and AFC. The proposed algorithm features two main modifications: (a) the Wiener variable step size (WVSS) and (b) the gradient spectral variance smoothing (GSVS). In AEC simulations, the WVSS-GSVS-FDAF-PEM- AFROW algorithm obtains outstanding robustness and smooth adaptation in highly adverse scenarios such as in bursting DT at high levels, and in a change of acoustic path during continuous DT. Similarly, in AFC simulations, the algorithm outperforms state-of-the-art algorithms when using a low-order near-end speech model and in colored non-stationary noise. © 2014 Elsevier B.V. ispartof: Signal Processing vol:104 pages:1-14 status: published

Acoustic echo cancellation (AEC) is used in speech communication systems where the existence of echoes degrades the speech intelligibility. Standard approaches to AEC rely on the assumption that the echo path to be identified can be modeled by a linear filter. However, some elements introduce nonlinear distortion and must be modeled as nonlinear systems. Several nonlinear models have been used with more or less success. The kernel affine projection algorithm (KAPA) has been successfully applied to many areas in signal processing but not yet to nonlinear AEC (NLAEC). The contribution of this paper is three-fold: (1) to apply KAPA to the NLAEC problem, (2) to develop a sliding-window leaky KAPA (SWL-KAPA) that is well suited for NLAEC applications, and (3) to propose a kernel function, consisting of a weighted sum of a linear and a Gaussian kernel. In our experiment set-up, the proposed SWL-KAPA for NLAEC consistently outperforms the linear APA, resulting in up to 12 dB of improvement in ERLE at a computational cost that is only 4.6 times higher. Moreover, it is shown that the SWL-KAPA outperforms, by 4-6 dB, a Volterra-based NLAEC, which itself has a much higher 413 times computational cost than the linear APA.

Recently, optimal linearly constrained minimum variance (LCMV) filtering methods have been applied to fundamental frequency estimation. Such estimators often yield preferable performance but suffer from being computationally cumbersome as the resulting cost functions are multimodal with narrow peaks and require matrix inversions for each point in the search grid. In this paper, we therefore consider fast implementations of LCMV-based fundamental frequency estimators, exploiting the estimators' inherently low displacement rank of the used Toeplitz-like data covariance matrices, using as such either the classic time domain averaging covariance matrix estimator, or, if aiming for an increased spectral resolution, the covariance matrix resulting from the application of the recent iterative adaptive approach (IAA). The proposed exact implementations reduce the required computational complexity with several orders of magnitude, but, as we show, further computational savings can be obtained by the adoption of an approximative IAA-based data covariance matrix estimator, reminiscent of the recently proposed Quasi-Newton IAA technique. Furthermore, it is shown how the considered pitch estimators can be efficiently updated when new observations become available. The resulting time-recursive updating can reduce the computational complexity even further. The experimental results show that the performances of the proposed methods are comparable or better than that of other competing methods in terms of spectral resolution. Finally, it is shown that the time-recursive implementations are able to track pitch fluctuations of synthetic as well as real-life signals. Recently, optimal linearly constrained minimum variance (LCMV) filtering methods have been applied to fundamental frequency estimation. Such estimators often yield preferable performance but suffer from being computationally cumbersome as the resulting cost functions are multimodal with narrow peaks and require matrix inversions for each point in the search grid. In this paper, we therefore consider fast implementations of LCMV-based fundamental frequency estimators, exploiting the estimators' inherently low displacement rank of the used Toeplitz-like data covariance matrices, using as such either the classic time domain averaging covariance matrix estimator, or, if aiming for an increased spectral resolution, the covariance matrix resulting from the application of the recent iterative adaptive approach (IAA). The proposed exact implementations reduce the required computational complexity with several orders of magnitude, but, as we show, further computational savings can be obtained by the adoption of an approximative IAA-based data covariance matrix estimator, reminiscent of the recently proposed Quasi-Newton IAA technique. Furthermore, it is shown how the considered pitch estimators can be efficiently updated when new observations become available. The resulting time-recursive updating can reduce the computational complexity even further. The experimental results show that the performances of the proposed methods are comparable or better than that of other competing methods in terms of spectral resolution. Finally, it is shown that the time-recursive implementations are able to track pitch fluctuations of synthetic as well as real-life signals.

In this paper, we propose a novel sequential error concealment algorithm for video and images based on sparse linear prediction. Block-based coding schemes in packet loss environments are considered. Images are modelled by means of linear prediction, and missing macroblocks are sequentially reconstructed using the available groups of pixels. The optimal predictor coefficients are computed by applying a missing data regression imputation procedure with a sparsity constraint. Moreover, an efficient procedure for the computation of these coefficients based on an exponential approximation is also proposed. Both techniques provide high-quality reconstructions and outperform the state-of-the-art algorithms both in terms of PSNR and MS-SSIM.

This paper presents weighted approaches for integrated active noise control and noise reduction in hearing aids. The unweighted integrated active noise control and noise reduction scheme introduced in the previous work does not allow to trade-off between the active noise control and the noise reduction. In some circumstances it will, however, be useful to emphasize one of the functional blocks. Changing the original optimisation problem to a constrained optimisation problem leads to a scheme based on a weighted mean squared error criterion that allows to focus either on the active noise control or on the noise reduction. It is similarly possible to derive a scheme that allows to focus either on reducing the speech distortion or on reducing the residual noise at the eardrum. In a single speech source scenario and when the number of sound sources (speech plus noise sources) is less than or equal to the number of microphones, it is possible to derive a simple formula for the output signal-to-noise ratio of the latter scheme. It can then be shown that this scheme delivers a constant signal-to-noise ratio at the eardrum for any weighting factor.

A probe noise signal can be used in an acoustic feedback cancellation system to prevent biased adaptive estimation of acoustic feedback paths. However, practical experiences and simulation results indicate that when-ever a low-level and inaudible probe noise signal is used, the convergence rate of the adaptive estimation is significantly decreased when keeping the steady-state error unchanged. The goal of this work is to derive analytic expressions for the system behavior such as convergence rate and steady-state error for a multiple-microphone and single-loudspeaker audio system, where the acoustic feedback cancellation is carried out using a probe noise signal. The derived results show how different system parameters and signal properties affect the cancellation performance, and the results explain theoretically the decreased convergence rate. Understanding this is important for making further improvements in the existing probe noise approach.

In this work, we analyze a general multiple-microphone and single-loudspeaker audio processing system, where a multichannel adaptive system is used to cancel the effect of acoustic feedback/echo, and a beamformer processes the feedback/echo canceled signals. We introduce and derive an accurate approximation of a frequency domain measure - the power transfer function - and show how it can be used to predict the convergence rate, system stability bound and the steady-state behavior of the entire cancellation system across frequency and time. We consider three example adaptive algorithms in the cancellation system: the least mean square, normalized least mean square, and the recursive least squares algorithms. Furthermore, we derive expressions to determine the step size parameter in the adaptive algorithms to achieve a desired system behavior, e.g., convergence rate at a specific frequency. Finally, we compare and discuss the performance of all three adaptive algorithms, and we verify the derived expressions through simulation experiments.

Multiple descriptions (MDs) is a method to obtain reliable signal transmissions on erasure channels. An MD encoder forms several descriptions of the signal and each description is independently transmitted across an erasure channel. The reconstruction quality then depends on the set of received descriptions. In this paper, we consider the design of redundant descriptions in an MD setup using $l_{1}$-minimization with Euclidean distortion constraints. In this way we are able to obtain sparse descriptions using convex optimization. The proposed method allows for an arbitrary number of descriptions and supports both symmetric and asymmetric distortion design. We show that MDs with partial overlapping information corresponds to enforcing coupled constraints in the proposed convex optimization problem. To handle the coupled constraints, we apply dual decompositions which makes first-order methods applicable and thereby admit solutions for large-scale problems, e.g., coding entire images or image sequences. We show by examples that the proposed framework generates non-trivial sparse descriptions and non-trivial refinements. We finally show that the sparse descriptions can be quantized and encoded using off-the-shell encoders such as the set partitioning in hierarchical trees (SPIHT) encoder, however, the proposed method shows a rate-distortion loss compared to state-of-the-art image MD encoders.

This paper analyses the output signal-to-noise ratio for a standard noise reduction scheme based on the multichannel Wiener filter and for an integrated active noise control and noise reduction scheme based on the filtered-X multichannel Wiener filter, both applied in a hearing aid framework that includes the effects of signal leakage through an open fitting and secondary path effects. In previous work, integrating noise reduction and active noise control has been shown to allow to compensate for effects of signal leakage and secondary path effects. These experimental results are now verified theoretically. The output signal-to-noise ratios are derived under a single speech source scenario. Theoretical results are then compared to simulations for a single noise source scenario and a multiple noise sources scenario.

In this paper, we present an iterative method for estimation of pitches from signals containing multiple sources using subspace techniques. The resulting estimator is termed Iterative Harmonic MUltiple SIgnal Classification (I-HMUSIC). Different modifications of I-HMUSIC are proposed that improve upon the classical MUSIC algorithm, including a computationally efficient method for noise subspace updating I-HMUSIC and its modifications are evaluated and compared with both the Cramer-Rao lower bound (CRLB) and non-iterative HMUSIC; good statistical performances have been obtained.

In this paper, we consider the problem of minimizing a non-smooth convex problem using first-order methods. The number of iterations required to guarantee a certain accuracy for such problems is often excessive and several methods, e.g., restart methods, have been proposed to speed-up the convergence. In the restart method a smoothness parameter is adjusted such that smoother approximations of the original non-smooth problem are solved in a sequence before the original, and the previous estimate is used as the starting point each time. Instead of adjusting the smoothness parameter after each restart, we propose a method where we modify the smoothness parameter in each iteration. We prove convergence and provide simulation examples for two typical signal processing applications, namely total variation denoising and l1-norm minimization. The simulations demonstrate that the proposed method require fewer iterations and show lower complexity compared to the restart method.

A multi-channel noise reduction technique is presented based on a Speech Distortion-Weighted Multi-channel Wiener Filter (SDW-MWF) approach that incorporates the conditional Speech Presence Probability (SPP). A traditional SDW-MWF uses a fixed parameter to a trade-off between noise reduction and speech distortion without taking speech presence into account. Consequently, the improvement in noise reduction comes at the cost of a higher speech distortion since the speech dominant segments and the noise dominant segments are weighted equally. Incorporating the conditional SPP in SDW-MWF allows to exploit the fact that speech may not be present at all frequencies and at all times, while the noise can indeed be continuously present. In speech dominant segments it is then desirable to have less noise reduction to avoid speech distortion, while in noise dominant segments it is desirable to have as much noise reduction as possible. Experimental results with hearing aid scenarios demonstrate that the proposed SDW-MWF incorporating the conditional SPP improves the signal-to-noise ratio compared to a traditional SDW-MWF.

We propose a new multiple description spherical quantization with repetitively coded amplitudes (MDSQRA) scheme suited for quantization of sinusoidal parameters. The quantization scheme is constituted by a set of spherical quantizers inspired by the multiple description spherical trellis-coded quantization (MDSTCQ) scheme. In this scheme, we apply repetitive coding on the amplitudes, while multiple description coding are applied on the phases and frequencies. Thereby, MDSQRA becomes directly implementable, as opposed to MDSTCQ, since the phase and frequency quantizers depend on the amplitudes which have dissimilar descriptions in MDSTCQ. Furthermore, we implement MDSQRA into a perceptual matching pursuit based sinusoidal audio coder. Finally, we evaluate MDSQRA through perceptual distortion measurements and MUSHRA listening tests. The tests show that MDSQRA outperforms MDSTCQ with respect to a expected perceptual distortion measure. The same results are obtained through the MUSHRA tests performed on sound clips coded using MDSQRA and MDSTCQ.

The problem of estimating frequencies of sinusoids in noise has been studied intensively by the signal processing community during the last decades. Traditionally high resolution subspace-based techniques suffer from high computational complexity, and generally sensitive to the colored noise. We present here a frequency-domain based subspace parameter estimation algorithm termed frequency-selective MUltiple SIgnal Classification (F-MUSIC) that is based on the signal and noise subspace orthogonality property. The method is computationally efficient in providing estimates in the selected subband compared to the classic MUSIC. The performance of F-MUSIC is evaluated and compared to both MUSIC and Cramer-Rao lower bound (CRLB). In a low signal to noise ratio (SNR) with colored noise scenarios, F-MUSIC outperforms MUSIC.

Recently, we proposed using Capon's minimum variance principle to find the fundamental frequency of a periodic waveform. The resulting estimator is formed such that it maximizes the output power of a bank of filters. We present an alternative optimal single filter design and then proceed to quantify the similarities and differences between the estimators using asymptotic analysis and Monte Carlo simulations. Our analysis shows that the single filter can be expressed in terms of the optimal filterbank and that the methods are asymptotically equivalent but generally different for finite length signals.

A new scheme for sinusoidal audio coding named multiple description spherical trellis-coded quantization is proposed and analytic expressions for the point densities and expected distortion of the quantizers are derived based on a high-resolution assumption. The proposed quantizers are of variable dimensions meaning that any number of sinusoids can be quantized jointly for each audio segment whereby a lower distortion is achieved compared to previously published scalar spherical quantizers. The quantizers are designed to minimize a perceptual distortion measure subject to an entropy constraint for a given packet-loss probability. In experiments, the performance of the quantizers is assessed and compared to the corresponding single description spherical quantizer and associated bounds under various conditions and is found to increase robustness towards packet-loss.

In this paper, we formulate the multi-pitch estimation problem and propose a number of methods to estimate the set of fundamental frequencies. The methods, which are based on nonlinear least-squares, multiple signal classification (MUSIC) and the Capon principles, have in common the fact that the multiple fundamental frequencies are estimated by means of a one-dimensional search. The statistical properties of the methods are evaluated via Monte Carlo simulations.

As voice over IP proliferates, packet loss concealment (PLC) at the receiver has emerged as an important factor in determining voice quality of service. Through the use of heuristic variations of signal and parameter repetition and overlap-add interpolation to handle packet loss, conventional PLC systems largely ignore the dynamics of the statistical evolution of the speech signal, possibly leading to perceptually annoying artifacts. To address this problem, we propose the use of hidden Markov models for PLC. With a hidden Markov model (HMM) tracking the evolution of speech signal parameters, we demonstrate how PLC is performed within a statistical signal processing framework. Moreover, we show how the HMM is used to index a specially designed PLC module for the particular signal context, leading to signal-contingent PLC. Simulation examples, objective tests, and subjective listening tests are provided showing the ability of an HMM-based PLC built with a sinusoidal analysis/synthesis model to provide better loss concealment than a conventional PLC based on the same sinusoidal model for all types of speech signals, including onsets and signal transitions.

In this letter, we present a decomposition for sinusoidal coding of audio, based on an amplitude modulation of sinusoids via a linear combination of arbitrary basis vectors. The proposed method, which incorporates a perceptual distortion measure, is based on a relaxation of a nonlinear least-squares minimization. Rate-distortion curves and listening tests show that, compared to a constant-amplitude sinusoidal coder, the proposed decomposition offers perceptually significant improvements in critical transient signals.

Psychoacoustical models have been used extensively within audio coding applications over the past decades. Recently, parametric coding techniques have been applied to general audio and this has created the need for a psychoacoustical model that is specifically suited for sinusoidal modelling of audio signals. In this paper, we present a new perceptual model that predicts masked thresholds for sinusoidal distortions. The model relies on signal detection theory and incorporates more recent insights about spectral and temporal integration in auditory masking. As a consequence, the model is able to predict the distortion detectability. In fact, the distortion detectability defines a (perceptually relevant) norm on the underlying signal space which is beneficial for optimisation algorithms such as rate-distortion optimisation or linear predictive coding. We evaluate the merits of the model by combining it with a sinusoidal extraction method and compare the results with those obtained with the ISO MPEG-1 Layer I-II recommended model. Listening tests show a clear preference for the new model. More specifically, the model presented here leads to a reduction of more than 20% in terms of number of sinusoids needed to represent signals at a given quality level.

The problem of modeling a signal segment as a sum of exponentially damped sinusoidal components is of interest in a wide range of fields, including speech and audio processing. Often, model parameters are estimated using subspace based techniques that exploit the so-called shift-invariance property. A drawback of these estimation techniques in relation to speech and audio processing is that the perceptual relevance of the model components is not taken into account. In this paper we show how to combine well-known subspace based estimation techniques with a recently developed perceptual distortion measure, to obtain an algorithm for extracting perceptually relevant model components. In analysis-synthesis experiments with wideband audio signals, objective and subjective evaluations show that the proposed algorithm improves perceived signal quality considerably over traditional subspace based analysis methods.

An extended sinusoidal model for speech signal processing is proposed. In this model, the frequencies of the sinusoidal components are constrained to be harmonically related, but their amplitudes are allowed to vary exponentially with time. Algorithms for model parameter estimation are derived. The proposed model shows considerable objective and subjective improvements in speech segments with transitions, especially in speech onsets, compared to the constant-frequency, constant-amplitude harmonic model. Potential application areas include speech synthesis and multimode speech coding.

A generalized sinusoidal model for speech signal processing is studied. The main feature of the model is that the amplitude of each sinusoidal component is allowed to vary exponentially with time. We propose to use the model in transitional speech segments such as speech onsets and voiced/unvoiced transitions. Computer simulations with natural speech signals indicate substantial better modeling performance in both transitional and voiced regions compared with the traditional constant-amplitude sinusoidal model.

Proposed is a frequency selective (FS) subspace-based method for determining the model order. A study is made of its performance when applied to estimating the number of sinusoids in white noise. Employing an FS-version of the ESPRIT algorithm, the recent ESTER model order estimation algorithm is extended to allow for the model order estimation on a frequency subset.

In this letter, we propose joint quantization of the parameters of a set of sinusoids based on the theory of trellis-coded quantization. A particular advantage of this approach is that it allows for joint quantization of a variable number of sinusoids, which is particularly relevant in variable rate parametric audio coding.Under high-resolution assumptions and based on a perceptually relevant distortion measure, we derive analytical expressions for the optimal design subject to an entropy constraint. Numerical experiments show a significant performance gain compared to optimal spherical quantization at the cost of a slight increase in computational complexity.