Near-end Speaker Research Articles

An online algorithm for echo cancellation, dereverberation and noise reduction based on a Kalman-EM Method

Many modern smart devices are equipped with a microphone array and a loudspeaker (or are able to connect to one). Acoustic echo cancellation algorithms, specifically their multi-microphone variants, are essential components in such devices. On top of acoustic echos, other commonly encountered interference sources in telecommunication systems are reverberation, which may deteriorate the desired speech quality in acoustic enclosures, specifically if the speaker distance from the array is large, and noise. Although sub-optimal, the common practice in such scenarios is to treat each problem separately. In the current contribution, we address a unified statistical model to simultaneously tackle the three problems. Specifically, we propose a recursive EM (REM) algorithm for solving echo cancellation, dereverberation and noise reduction. The proposed approach is derived in the short-time Fourier transform (STFT) domain, with time-domain filtering approximated by the convolutive transfer function (CTF) model. In the E-step, a Kalman filter is applied to estimate the near-end speaker, based on the noisy and reveberant microphone signals and the echo reference signal. In the M-step, the model parameters, including the acoustic systems, are inferred. Experiments with human speakers were carried out to examine the performance in dynamic scenarios, including a walking speaker and a moving microphone array. The results demonstrate the efficiency of the echo canceller in adverse conditions together with a significant reduction in reverberation and noise. Moreover, the tracking capabilities of the proposed algorithm were shown to outperform baseline methods.

An Improved Double-talk Detection Algorithm for Echo Cancellation in Teleconference System

Echo cancellation commonly employs a double-talk detector (DTD), which is essential to keep the adaptive filter from diverging in the presence of near-end speech. An improved double-talk detection algorithm is proposed to promote the performance of the echo cancellation in teleconference system in the presence of speech issued from the near-end speaker (double-talk). Based on the partitioned block frequency domain adaptive filter algorithm (PBFDAF), the advanced double-talk detection method is presented in this paper constructs a joint correlation coefficient variable to judge the communication state and the variable is applied as a smoothing parameter to the error signal and then get the final output signal. Simulation results show that the improved algorithm can effectively detect the current voice state, eliminate the echo signal, and keep the near-end speaker’s voice well. The performance of the echo cancellation algorithm is improved when both echo and near end signals are simultaneously active by the proposed technique.

Model-Based Method for Acoustic Echo Cancelation and Near-End Speaker Extraction: Non-negative Matrix Factorization

Rapid escalation of wireless communication and hands-free telephony creates a problem with acoustic echo in full-duplex communication applications. In this paper a simulation of model-based acoustic echo cancelation and near-end speaker extraction using statistical methods relying on nonnegative matrix factorization (NMF) is proposed. Acoustic echo cancelation using the NMF algorithm is developed and its implementation is presented, along with all positive, real time elements and factorization techniques. Experimental results are compared against the widely used existing adaptive algorithms which have a disadvantage in terms of long impulse response, increased computational load and wrong convergence due to change in near-end enclosure. All these shortcomings have been eliminated in the statistical method of NMF that reduces echo and enhances audio signal processing.

Array model interpolation and subband iterative adaptive filters applied to beamforming-based acoustic echo cancellation.

In this paper, an evolutionary exposition is given in regard to the enhancing strategies for acoustic echo cancellers (AECs). A fixed beamformer (FBF) is utilized to focus on the near-end speaker while suppressing the echo from the far end. In reality, the array steering vector could differ considerably from the ideal freefield plane wave model. Therefore, an experimental procedure is developed to interpolate a practical array model from the measured frequency responses. Subband (SB) filtering with polyphase implementation is exploited to accelerate the cancellation process. Generalized sidelobe canceller (GSC) composed of an FBF and an adaptive blocking module is combined with AEC to maximize cancellation performance. Another enhancement is an internal iteration (IIT) procedure that enables efficient convergence in the adaptive SB filters within a sample time. Objective tests in terms of echo return loss enhancement (ERLE), perceptual evaluation of speech quality (PESQ), word recognition rate for automatic speech recognition (ASR), and subjective listening tests are conducted to validate the proposed AEC approaches. The results show that the GSC-SB-AEC-IIT approach has attained the highest ERLE without speech quality degradation, even in double-talk scenarios.

A Signal Decorrelation PNLMS Algorithm for Double-Talk Acoustic Echo Cancellation

Acoustic echo cancellation (AEC) in voiced communication systems is used to eliminate the echo which corrupts the speech signal and reduces the efficiency of signal transmission. Usually, the performance of AEC system based on the adaptive filtering degrades seriously in the presence of speech issued from the near-end speaker (double-talk). In typical AEC scenarios, double-talk detector (DTD) must be added to AEC for improving speech quality. One of the main problems in AEC with DTD is that the DTD errors can result in either large residual echo or distorting the near-end input speech. Considering the strong correlation property of speech signals, this paper presents a novel proportionate decorrelation normalized least-mean-square (PDNLMS) adaptive AEC without DTD for echo cancellation as an interesting alternative to the typical AEC with DTDs. Unlike traditional AEC with a DTD, the proposed PDNLMS uses the difference of near-end speech as the residual error to update adaptive echo channel filter during the periods of double-talk, which can efficiently reduce the double-talk influence on the AEC adaptation process. The experimental results show that not only the proposed PDNLMS without DTD illustrate better stability and faster convergence rate, but it is also of a lower steady-state misalignment and better residual signal than current methods with DTDs at a lower computational cost.

Adaptive beamforming for acoustic echo cancellation using measured array models and subband filtering

Acoustic echo that can substantially undermine speech quality is one of the key issues must be addressed in practical telecommunication systems. In this paper, an evolutionary exposition is given in regard to the enhancing strategies for acoustic echo cancelers (AEC). On the basis basic adaptive filtering, a highly directional microphone array is designed as a fixed beamformer (FBF) to focus on the near-end speaker and suppress the echo from the far-end as well as noise and interference from the background. Design of the beamformer is based on array models interpolated from measured frequency responses. Subband (SB) filtering with polyphase decomposition is exploited to accelerate the cancellation process with multiple choice of step size in the subbands. To further enhance the canceler, adaptive generalized sidelobe canceler (GSC) can be utilized, which is comprised of a FBF for the near-end speech and an adaptive blocking module for the echo and noise. Objective tests in terms of echo return loss enhancement (ERLE) and perceptual evaluation of speech quality (PESQ) are conducted to compare four algorithms with various combinations of enhancement. The results show that the GSC-SB-AEC approach has attained the highest ERLE and best speech quality, even in a double-talk scenario.

Wavelet-Based ICA Using Maximum Likelihood Estimation and Information-Theoretic Measure for Acoustic Echo Cancellation During Double Talk Situation

Acoustic echo cancellation (AEC) plays a prominent role in the present-day hands-free communication environment, owing to the usage of adaptive digital filter techniques. In a duplex communication scenario, there is a need for a double talk detection algorithm in the near-end speaker system which disables the update of the adaptive filter coefficients, thus hindering the process of echo cancellation and leading to a partial solution. To completely solve this problem, independent component analysis (ICA) is used to separate the far-end echo from the mixture of the near-end speech and the far-end echo signal. This paper proposes a new adaptive digital filter using maximum likelihood estimation of ICA and minimization of mutual information of ICA techniques for AEC. The advancement to this technique is made by transforming the observation into an adequate representation using wavelet decomposition. The performance of the echo cancellation is measured in terms of the echo return loss enhancement (ERLE). Higher ERLE indicates better echo cancellation. From the simulation results, it is found that the minimization of the mutual information of ICA has a higher value of ERLE than that by maximum likelihood estimation. The efficiency of the system thus increases by minimizing the processing time using wavelet ICA-based adaptive filter over the conventional adaptive filter.

A Single-Channel MVDR Filter for Acoustic Echo Suppression

Acoustic echo suppression techniques for full-duplex communication in low-complexity systems are commonly known to introduce distortion to the desired signal (i.e., near-end speech). Moreover, most traditional echo control techniques typically require accurately detecting the contribution of the near-end speaker to the microphone signal (“double talk”). In this letter, we propose a novel approach to acoustic echo suppression, which aims at extracting the near-end signal using a constraint for minimizing the distortion, and without requiring a double-talk detector.

Echo Cancellation—A Likelihood Ratio Test for Double-Talk Versus Channel Change

Echo cancellers (ECs) are in wide use in both electrical (four-wire to two-wire mismatch) and acoustic (speaker-microphone coupling) applications. One of the main design problems is the control logic for adaptation. Basically, the algorithm weights should be frozen in the presence of double-talk and adapt quickly in the absence of double-talk. The control logic can be quite complicated since it is often not easy to discriminate between the echo signal and the near-end speaker. This paper derives a log-likelihood ratio test (LRT) for deciding between double-talk (freeze weights) and a channel change (adapt quickly) using a stationary Gaussian stochastic input signal model. The probability density function (pdf) of a sufficient statistic under each hypothesis is obtained, and the performance of the test is evaluated as a function of the system parameters. The receiver operating characteristics (ROCs) indicate that it is difficult to correctly decide between double-talk and a channel change based upon a single look. However, postdetection integration of approximately 100 sufficient statistic samples yields a detection probability close to unity (0.99) with a small false-alarm probability (0.01)

Control of an adaptive echo canceller using a near-end speech detector

In the context of hands-free mobile telephony, acoustic echo cancellation (AEC) remains an important problem to deal with. This work addresses the study of an acoustic echo canceller controlled by a near-end speech detector. As a matter of fact, the echo canceller is an adaptive filter whose coefficients are disturbed by the presence of speech coming from the near-end speaker. Consequently, it is useful to stop the adaptation when this signal is detected to get a better estimation of the echo (far-end speaker). In this paper, a near-end speech detector based on the partial coherence – computed from the observations received by two microphones conditioned on the signal emitted by the loudspeaker – is investigated. To improve the performance of the AEC system, two approaches performing a saving of the AEC coefficients are proposed. The system is assessed through the power of the residual echo.

Recursive system identification in the presence of burst disturbance

Conventional algorithms for the recursive identification of time varying systems will fail in the presence of disturbances which occur in bursts. Recently proposed methods to solve the problem suffer from certain drawbacks. In this paper, we present a new algorithm for FIR system identification which automatically detects the onset of a burst of disturbance and inhibits the updating of the parameters for the duration of the burst. Parameter updating is continued after the burst ceases, enabling the tracking of slow variations in the system parameters. Simulations are used to study the performance of the algorithm in the context of the well-known problem of near-end speaker interference in echo cancellation. The results demonstrate that the algorithm performs as predicted.