Multichannel Wiener Filter Research Articles

Multichannel Wiener Filters in Binaural and Bilateral Hearing Aids — Speech Intelligibility Improvement and Robustness to DoA Errors

Multichannel Wiener Filters in Binaural and Bilateral Hearing Aids — Speech Intelligibility Improvement and Robustness to DoA Errors

Non-Stationary Noise PSD Matrix Estimation for Multichannel Blind Speech Extraction

Noise power spectral density (PSD) matrix estimation is one of the most important components of a multichannel blind speech extraction framework, as it largely determines the amount of residual noise at the output of a spatial filter. Optimality of well-known spatial filters, such as the multichannel Wiener filter, is only ensured if the PSD matrices of the noise and the desired speech are accurately estimated. In practical situations, where the noise is nonstationary, temporal averaging over time frames where the desired signal is inactive does not provide sufficiently fast tracking of the noise PSD matrix, resulting in high residual noise at the spatial filter output. Therefore, approaches that estimate the PSD matrices using narrowband signal detection have been proposed. Following the well-known single- and multichannel minima-controlled recursive averaging (MCRA) approaches, in this paper, we focus on narrowband speech presence probability-based noise PSD matrix estimators, which are suitable for blind scenarios where the location and the propagation vector of the desired speech source are unknown. The main contributions of the paper are a maximum likelihood interpretation of the multichannel MCRA, and a coherent-to-diffuse ratio-based a priori speech absence probability (SAP) estimator. The latter is a key parameter that determines the accuracy of the noise PSD matrix estimates in nonstationary scenarios. In this paper, we confirm the importance of the a priori SAP and show that its control is crucial for source extraction in nonstationary environments.

Adaptive attention-driven speech enhancement for EEG-informed hearing prostheses.

State-of-the-art hearing prostheses are equipped with acoustic noise reduction algorithms to improve speech intelligibility. Currently, one of the major challenges is to perform acoustic noise reduction in so-called cocktail party scenarios with multiple speakers, in particular because it is difficult-if not impossible-for the algorithm to determine which are the target speaker(s) that should be enhanced, and which speaker(s) should be treated as interfering sources. Recently, it has been shown that electroencephalography (EEG) can be used to perform auditory attention detection, i.e., to detect to which speaker a subject is attending based on recordings of neural activity. In this paper, we combine such an EEG-based auditory attention detection (AAD) paradigm with an acoustic noise reduction algorithm based on the multi-channel Wiener filter (MWF), leading to a neuro-steered MWF. In particular, we analyze how the AAD accuracy affects the noise suppression performance of an adaptive MWF in a sliding-window implementation, where the user switches his attention between two speakers.

Phase reference for the generalized multichannel Wiener filter

The multichannel Wiener filter (MWF) is a well-established noise reduction technique for speech processing. Most commonly, the speech component in a selected reference microphone is estimated. The choice of this reference microphone influences the broadband output signal-to-noise ratio (SNR) as well as the speech distortion. Recently, a generalized formulation for the MWF (G-MWF) was proposed that uses a weighted sum of the individual transfer functions from the speaker to the microphones to form a better speech reference resulting in an improved broadband output SNR. For the MWF, the influence of the phase reference is often neglected, because it has no impact on the narrow-band output SNR. The G-MWF allows an arbitrary choice of the phase reference especially in the context of spatially distributed microphones. In this work, we demonstrate that the phase reference determines the overall transfer function and hence has an impact on both the speech distortion and the broadband output SNR. We propose two speech references that achieve a better signal-to-reverberation ratio (SRR) and an improvement in the broadband output SNR. Both proposed references are based on the phase of a delay-and-sum beamformer. Hence, the time-difference-of-arrival (TDOA) of the speech source is required to align the signals. The different techniques are compared in terms of SRR and SNR performance.

Maximum Likelihood PSD Estimation for Speech Enhancement in Reverberation and Noise

In this contribution, we focus on the problem of power spectral density (PSD) estimation from multiple microphone signals in reverberant and noisy environments. The PSD estimation method proposed in this paper is based on the maximum likelihood (ML) methodology. In particular, we derive a novel ML PSD estimation scheme that is suitable for sound scenes which besides speech and reverberation consists of an additional noise component whose second-order statistics are known. The proposed algorithm is shown to outperform an existing similar algorithm in terms of PSD estimation accuracy. Moreover, it is shown numerically that the mean-squared estimation error achieved by the proposed method is near the limit set by the corresponding Cramer--Rao lower bound. The speech dereverberation performance of a multichannel Wiener filter based on the proposed PSD estimators is measured using several instrumental measures and is shown to be higher than when the competing estimator is used. Moreover, we perform a speech intelligibility test where we demonstrate that both the proposed and the competing PSD estimators lead to similar intelligibility improvements.

Multichannel Audio Source Separation With Deep Neural Networks

This article addresses the problem of multichannel audio source separation. We propose a framework where deep neural networks (DNNs) are used to model the source spectra and combined with the classical multichannel Gaussian model to exploit the spatial information. The parameters are estimated in an iterative expectation-maximization (EM) fashion and used to derive a multichannel Wiener filter. We present an extensive experimental study to show the impact of different design choices on the performance of the proposed technique. We consider different cost functions for the training of DNNs, namely the probabilistically motivated Itakura–Saito divergence, and also Kullback–Leibler, Cauchy, mean squared error, and phase-sensitive cost functions. We also study the number of EM iterations and the use of multiple DNNs, where each DNN aims to improve the spectra estimated by the preceding EM iteration. Finally, we present its application to a speech enhancement problem. The experimental results show the benefit of the proposed multichannel approach over a single-channel DNN-based approach and the conventional multichannel nonnegative matrix factorization-based iterative EM algorithm.

GEVD-Based Low-Rank Approximation for Distributed Adaptive Node-Specific Signal Estimation in Wireless Sensor Networks

In this paper, we address the problem of distributed adaptive estimation of node-specific signals for signal enhancement or noise reduction in wireless sensor networks with multi-sensor nodes. The estimation is performed by a multi-channel Wiener filter (MWF) in which a low-rank approximation based on a generalized eigenvalue decomposition (GEVD) is incorporated. In non-stationary or low-SNR conditions, this GEVD-based MWF has been demonstrated to be more robust than the original MWF. In a centralized realization where a fusion center has access to all the nodes’ sensor signal observations, the network-wide sensor signal correlation matrices and the low-rank approximation can be directly estimated and used to compute the network-wide GEVD-based MWF. However, in this paper, we aim to avoid centralizing the sensor signal observations, in which case the network-wide sensor signal correlation matrices cannot be estimated. To this end, we start from the so-called distributed adaptive node-specific signal estimation (DANSE) algorithm, and include GEVD-based low-rank approximations in the per-node local computations. Remarkably, the new algorithm is able to significantly compress the signal observations transmitted between the nodes, while still converging to the network-wide GEVD-based MWF as if each node would have access to all sensor signal observations, even though the low-rank approximations are applied locally at each node. We provide a theoretical convergence analysis, which shows that the algorithm converges to the network-wide GEVD-based MWF under conditions that are less strict than in the original DANSE algorithm. The convergence and performance of the algorithm are further investigated via numerical simulations.

Binaural Noise Cue Preservation in a Binaural Noise Reduction System With a Remote Microphone Signal

A general binaural noise reduction system is considered that employs the multichannel Wiener filter with partial noise estimation ( $\text{MWF}_\eta$ ) allowing for an explicit tradeoff between noise reduction and binaural noise cue preservation. In this paper, it is assumed that along with the general binaural system, a remote microphone signal with a high input signal-to-noise ratio (SNR) is available for inclusion in the $\text{MWF}_\eta$ . The use of this remote microphone signal with a high input SNR allows for a simultaneous increase in both noise reduction performance and preservation of the binaural noise cues. To further increase the performance, a modification to the partial noise estimation (PNE) variable, $\eta$ , is proposed which relies on exploiting the aforementioned trade-off by either constraining the output SNR or binaural noise cues to the same level before and after the addition of the remote microphone signal. The validity of the theoretical results are supplemented via simulations using a binaural setup with a single speech and noise source.

Fundamental Frequency Informed Speech Enhancement in a Flexible Statistical Framework

Conventional statistical clean speech estimators, like the Wiener filter, are frequently used for the spectro-temporal enhancement of noise corrupted speech. Most of these approaches estimate the clean speech independently for each time-frequency point, neglecting the structure of the underlying speech sound. In this work, we derive a statistical estimator that explicitly takes into account information about the characteristic structure of voiced speech by means of a harmonic signal model. To this end, we also present a way to estimate a harmonic model-based clean speech representation and the corresponding error variance directly in the short-time Fourier transform domain. The resulting estimator is optimal in the minimum-mean-squared error sense and can conveniently be formulated in terms of a multichannel Wiener filter. The proposed estimator outperforms several reference algorithms in terms of speech quality and intelligibility as predicted by instrumental measures.

Joint Dereverberation and Noise Reduction Based on Acoustic Multi-Channel Equalization

Regularized acoustic multi-channel equalization techniques, such as regularized partial multi-channel equalization based on the multiple-input/output inverse theorem (RPMINT), are able to achieve a high dereverberation performance in the presence of room impulse response perturbations but may lead to amplification of the additive noise. In this paper, two time-domain techniques aiming at joint dereverberation and noise reduction based on acoustic multi-channel equalization are proposed. The first technique, namely RPMINT for joint dereverberation and noise reduction (RPM-DNR), extends RPMINT by explicitly taking the noise statistics into account. In addition to the regularization parameter used in RPMINT, the RPM-DNR technique introduces an additional weighting parameter, enabling a trade-off between dereverberation and noise reduction. The second technique, namely multi-channel Wiener filter for joint dereverberation and noise reduction (MWF-DNR), takes both the speech and the noise statistics into account and uses the RPMINT filter to compute a dereverberated reference signal for the multi-channel Wiener filter. The MWF-DNR technique also introduces an additional weighting parameter, which now provides a trade-off between speech distortion and noise reduction. To automatically select the regularization and weighting parameters, for the RPM-DNR technique a novel procedure based on the L-hypersurface is proposed, whereas for the MWF-DNR technique two decoupled optimization procedures based on the L-curve are used. Extensive simulations demonstrate using instrumental measures that the RPM-DNR technique maintains the dereverberation performance of the RPMINT technique while improving its noise reduction performance. Furthermore, it is shown that the MWF-DNR technique yields a significantly better noise reduction performance than the RPM-DNR technique at the expense of a worse dereverberation performance.

Neuro-steered noise suppression for auditory prostheses

In a multi-speaker scenario, a major challenge for noise suppression systems in hearing instruments is to determine which sound source the listener is attending to. It has been shown that a linear decoder can extract a neural signal from EEG recordings that is better correlated with the envelope of the attended speech signal than with the envelopes of the other signals. This can be exploited to perform auditory attention detection (AAD), which can then steer a noise suppression algorithm. The speech signal is passed through a model of the auditory periphery before extracting its envelope. We compared 7 different periphery models and found that best AAD performance was obtained with a gamma-tone filter bank followed by power-law compression. Most AAD studies so far have employed a dichotic paradigm, wherein each ear receives a separate speech stream. We compared this to a more realistic setup where speech was simulated to originate form two different spatial locations, and found that although listening conditions were harder, AAD performance was better than for the dichotic setup. Finally, we designed a neuro-steered denoising algorithm that informs the voice activity detection stage of a multi-channel Wiener filter based on AAD, and found a large signal-to-noise-ratio improvement at the output.

Interaural Coherence Preservation in Multi-Channel Wiener Filtering-Based Noise Reduction for Binaural Hearing Aids

Besides noise reduction an important objective of binaural speech enhancement algorithms is the preservation of the binaural cues of all sound sources. To this end, an extension of the binaural multi-channel Wiener filter (MWF), namely the MWF-ITF, has been proposed, which aims to preserve the Interaural Transfer Function (ITF) of the noise sources. However, the MWF-ITF is well-suited only for directional noise sources but not for, e.g., spatially isotropic noise, whose spatial characteristics cannot be properly described by the ITF but rather by the Interaural Coherence (IC). Hence, another extension of the binaural MWF, namely the MWF-IC, has been recently proposed, which aims to preserve the IC of the noise component. Since for the MWF-IC a substantial tradeoff between noise reduction and IC preservation exists, in this paper we propose a perceptually constrained version of the MWF-IC, where the amount of IC preservation is controlled based on the IC discrimination ability of the human auditory system. In addition, a theoretical analysis of the binaural cue preservation capabilities of the binaural MWF and the MWF-ITF for spatially isotropic noise fields is provided. Several simulations in diffuse noise scenarios show that the perceptually constrained MWF-IC yields a controllable preservation of the IC without significantly degrading the output SNR compared to the binaural MWF and the MWF-ITF. Furthermore, contrary to the binaural MWF and MWF-ITF, the proposed algorithm retains the spatial separation between the output speech and noise components while the binaural cues of the speech component are only slightly distorted, such that the binaural hearing advantage for speech intelligibility can still be exploited.

Theoretical Analysis of Linearly Constrained Multi-Channel Wiener Filtering Algorithms for Combined Noise Reduction and Binaural Cue Preservation in Binaural Hearing Aids

Besides noise reduction, an important objective of binaural speech enhancement algorithms is the preservation of the binaural cues of all sound sources. For the desired speech source and the interfering sources, e.g., competing speakers, this can be achieved by preserving their relative transfer functions (RTFs). It has been shown that the binaural multi-channel Wiener filter (MWF) preserves the RTF of the desired speech source, but typically distorts the RTF of the interfering sources. To this end, in this paper we propose two extensions of the binaural MWF, i.e., the binaural MWF with RTF preservation (MWF-RTF) aiming to preserve the RTF of the interfering source and the binaural MWF with interference rejection (MWF-IR) aiming to completely suppress the interfering source. Analytical expressions for the performance of the binaural MWF, MWF-RTF and MWF-IR in terms of noise reduction, speech distortion and binaural cue preservation are derived, showing that the proposed extensions yield a better performance in terms of the signal-to-interference ratio and preservation of the binaural cues of the directional interference, while the overall noise reduction performance is degraded compared to the binaural MWF. Simulation results using binaural behind-the-ear impulse responses measured in a reverberant environment validate the derived analytical expressions for the theoretically achievable performance of the binaural MWF, MWF-RTF, and MWF-IR, showing that the performance highly depends on the position of the interfering source and the number of microphones. Furthermore, the simulation results show that the MWF-RTF yields a very similar overall noise reduction performance as the binaural MWF, while preserving the binaural cues of both the speech and the interfering source.

An evaluation of EEG ocular artifact removal with a multi-channel wiener filter based on probabilistic generative model.

Data contamination by ocular artifacts such as eye blinks and eye movements is a major barrier that must be overcome when attempting to analyze electroencephalogram (EEG) and event-related potential (ERP) data. To handle this problem, a number of artifact removal methods has been proposed. Specifically, we focus on a method using a multi-channel Wiener filters based on a probabilistic generative model. This method assumes that the observed signal is the sum of multiple signals elicited by psychological or physical events, and separates the observed signal into each event signal using estimated model parameters. Based on this scheme, we have proposed a model parameter estimation method using prior information of each event signal. In this paper, we examine the potential of this model to deal with highly contaminated signals by collecting EEG data intentionally contaminated by eye blinks and relatively clean ERP data, and using them as prior information of each event signal. We conducted an experimental evaluation using a classical attention task. The results showed the proposed method effectively enhances the target ERP component while reducing the contamination caused by eye blinks.

Suppression of noise and late reverberation based on blind signal extraction and Wiener filtering

We introduce a new optimized microphone-array processing method for a spoken-dialogue robot in noisy and reverberant environments. The method is based on frequency-domain blind signal extraction, a signal separation algorithm that exploits the sparseness of a speech signal to separate the target speech and diffuse background noise from the sound mixture captured by a microphone array. This algorithm is combined with multichannel Wiener filtering so that it can effectively suppress both background noise and reverberation, given a priori information of room reverberation time. In this paper, first, we develop an automatic optimization scheme based on the assessment of musical noise via higher-order statistics and acoustic model likelihood. Next, to maintain the optimum performance of the system, we propose the multimodal switching scheme using the distance information provided by robot's image sensor and the estimation of SNR condition. Experimental evaluations have been conducted to confirm the efficacy of this method.

Speech distortion weighted multi-channel Wiener filter and its application to speech recognition

Speech distortion weighted multi-channel Wiener filter and its application to speech recognition

Effective Binaural Multi-Channel Processing Algorithm for Improved Environmental Presence

Binaural noise-reduction algorithms based on multi-channel Wiener filter (MWF) are promising techniques to be used in binaural assistive listening devices. The real-time implementation of the existing binaural MWF methods, however, involves challenges to increase the amount of noise reduction without imposing speech distortion, and at the same time preserving the binaural cues of both speech and noise components. Although significant efforts have been made in the literature, most developed methods so far have focused only on either the former or latter problem. This paper proposes an alternative binaural MWF algorithm that incorporates the non-stationarity of the signal components into the framework. The main objective is to design an algorithm that would be able to select the sources that are present in the environment. To achieve this, a modified speech presence probability (SPP) and a single-channel speech enhancement algorithm are utilized in the formulation. The resulting optimal filter also avoids the poor estimation of the second-order clean speech statistics, which is normally done by simple subtraction. Theoretical analysis and performance evaluation using realistic recorded data shows the advantage of the proposed method over the reference MWF solution in terms of the binaural cues preservation, as well as the noise reduction and speech distortion.

Analysis of the average performance of the multi-channel Wiener filter for distributed microphone arrays using statistical room acoustics

For most multi-microphone noise reduction algorithms, e.g. the multi-channel Wiener filter (MWF), it is well known that the performance depends on the acoustic scenario at hand, i.e. the used microphone array, the position of the desired source and the noise field. Since the position of the desired source is not always known a priori, it is of great interest in many applications to be able to compute the average performance for a specific microphone array, which can be obtained by averaging the performance over all feasible source positions. A possible but either time-consuming or computationally complex approach to achieve this is to use measurements or simulations for a large number of source positions.In this paper, we propose to use the statistical properties of the acoustical transfer functions (ATFs) between the desired source and the microphones to derive analytical expressions for the spatially averaged performance measures (output SNR, noise reduction, speech distortion) of the MWF, assuming a homogeneous and known noise field. In addition, we show that although the spatially averaged performance measures do not express the performance of the MWF for a given position of the source and/or the microphones, they can be used to derive approximate analytical expressions for the average performance of the MWF for a given position of the microphones. Experimental results show that the proposed analytical expressions can be used to easily compare the performance of different microphone arrays, e.g. in an acoustic sensor network, without having to measure or numerically simulate a large number of ATFs.

Informed Spatial Filtering for Sound Extraction Using Distributed Microphone Arrays

Hands-free acquisition of speech is required in many human-machine interfaces and communication systems. The signals received by integrated microphones contain a desired speech signal, spatially coherent interfering signals, and background noise. In order to enhance the desired speech signal, state-of-the-art techniques apply data-dependent spatial filters which require the second order statistics (SOS) of the desired signal, the interfering signals and the background noise. As the number of sources and the reverberation time increase, the estimation accuracy of the SOS deteriorates, often resulting in insufficient noise and interference reduction. In this paper, a signal extraction framework with distributed microphone arrays is developed. An expectation maximization (EM)-based algorithm detects the number of coherent speech sources and estimates source clusters using time-frequency (TF) bin-wise position estimates. Subsequently, the second order statistics (SOS) are estimated using bin-wise speech presence probability (SPP) and a source probability for each source. Finally, a desired source is extracted using a minimum variance distortionless response (MVDR) filter, a multichannel Wiener filter (MWF) and a parametric multichannel Wiener filter (PMWF). The same framework can be employed for source separation, where a spatial filter is computed for each source considering the remaining sources as interferers. Evaluation using simulated and measured data demonstrates the effectiveness of the framework in estimating the number of sources, clustering, signal enhancement, and source separation.

Minimum Mutual Information-Based Linearly Constrained Broadband Signal Extraction

In this contribution, the problem of broadband acoustic signal extraction is treated as a specific source separation problem, where the desired signal components are to be separated from all remaining undesired components. For this, we exploit the generic TRIple-N Independent component analysis for CONvolutive mixtures (TRINICON) framework. The TRINICON optimization criterion is complemented with linear constraints leading to the Linearly Constrained Minimum Mutual Information (LCMMI) criterion for desired signal extraction. A general linearly constrained update rule for iterative filter optimization is derived, which can efficiently be realized in a novel Minimum Mutual Information (MMI)-Generalized Sidelobe Canceler (GSC). The general treatment of the signal extraction problem using an MMI criterion provides several advantages: Firstly, new insights into the signal extraction problem can be derived by establishing links to both the original GSC and the Multichannel Wiener Filter (MWF). Secondly, by exploiting fundamental properties characteristic for speech and audio signals, complicated and often unreliable Voice Activity Detection (VAD)-based control mechanisms become unnecessary. Thirdly, the overall realization requires only prior information of the desired source position. An evaluation of the MMI-GSC for the double-talk situation with two concurrently active speech sources under reverberant and noisy conditions demonstrates the effectiveness of this novel approach.