基于鲁棒波束形成的麦克风阵列系统

为了提高噪声消除的鲁棒性,传统的通用旁瓣消除器采用固定波束形成作为输出的系数约束自适应滤波器,采用CCAF的输出作为输入的标准约束自适应滤波器.此系统的跟踪性能随信号频率变化而改变,并且由于波束带宽的影响使得目标语音在低频发生畸变.采用优化波束形成方法来得到较宽的带宽,对输出进行后滤波处理来消除残余噪声.试验表明本方法可明显提高语音信噪比.     

麦克风阵列鲁棒频率不变波束形成算法

在实际应用中,频率不变波束形成器通常受到麦克风阵列失配误差的影响,因此提高频率不变波束形成器的鲁棒性具有重要意义。针对上述问题提出了一种约束优化模型,可以在保持频率不变波束形成的同时提高阵列的鲁棒性。首先设计目标波束图,考虑到差分麦克风阵列本身具有频率不变的波束图,选用传统二阶超心型差分麦克风波束图做为目标波束图。上述模型以麦克风阵列权矢量的二范数作为目标函数来最大化鲁棒性,在无失真约束,目标波束主瓣逼近约束以及旁瓣增益精准控制约束下实现频率不变。然后在交替方向乘子法算法框架下,将优化问题分解为多个优化子问题求解,然后对每个优化子问题分别求解,通过仿真验证了在交替方向乘子法算法下上述模型的可行性与有效性,最终达到了麦克风阵列鲁棒频率不变波束响应的效果。     

圆形平面麦克风阵列构型的优化设计方法

提出利用遗传算法来优化设计圆形麦克风阵列构型。构造了兼顾主瓣宽度和最大旁瓣增益的目标函数,使得两者之间有着较好折中。考虑到实用性,将圆形面板划分为1 cm×1 cm的方格,坐标只能在方格中心取值,确保优化阵列构型是可实现的。设计了选择、交叉和变异等操作的解决方案,保证整个算法可行。仿真结果表明:优化算法可以得到合适的阵列构型,相对于规则型阵列,这种优化阵列构型在主瓣宽度和最大旁瓣增益控制方面有着明显优势。     

麦克风阵列声源定位与跟踪性能改进

针对将线形全向麦克风阵列应用于室内声源定位与跟踪时出现的对称双主瓣、空间角分辨率下降、选择性频率衰落等问题,提出了相应的解决方法。联合使用所提出的三个方法,将大幅度提高室内线形全向麦克风阵列声源定位与跟踪的性能,而不必改变阵列拓扑结构、不增加硬件成本。     

麦克风阵列及其消噪性能研究

用麦克风阵列进行语音处理的方法可以提高信噪比,解决环境噪声,回声和混响引起的语音识别性能降低的问题,麦克风阵列系统是由一组按一定几何结构摆放的麦克组成的系统。此阵列能接收空间传播信号,经过适当的信号处理,提取所需的信号源和信号属性等信息。使用该系统可大大提高强干扰环境下的语音识别性能。     

面向多说话人分离的深度学习麦克风阵列语音增强

随着近年来人机语音交互场景不断增加,利用麦克风阵列语音增强提高语音质量成为研究热点之一。与环境噪声不同,多说话人分离场景下干扰说话人语音与目标说话人同为语音信号,呈现类似的时、频特性,对传统麦克风阵列语音增强技术提出更高的挑战。针对多说话人分离场景,基于深度学习网络构建麦阵空间响应代价函数并进行优化,通过深度学习模型训练设计麦克风阵列期望空间传输特性,从而通过改善波束指向性能提高分离效果。仿真和实验结果表明,该方法有效提高了多说话人分离性能。     

基于麦克风阵列的语音增强算法

为解决现有语音增强算法需要麦克风数量较多和受估计误差影响较大的问题,提出一种改进的声源定位和波束形成方法。在现有声源定位算法利用时间延迟的基础上,引入能量衰减参数,实现利用双麦克风进行声源定位的目标;在波束形成算法中引入加载系数,在出现协方差矩阵统计失配时仍可对期望方向聚焦,提高波束形成算法的鲁棒性。仿真结果表明,改进后的算法与传统算法相比具有更强的鲁棒性。     

麦克风阵列语音增强的研究

文章介绍了各种基本的麦克风阵列语音增强算法,对其消噪性能进行了系统地分析,并以实测数据进行了测试。并介绍了基于稳健波束形成器、近场超定向波束形成器、广义奇异值分解和传输函数广义旁瓣相消器等结构的麦克风阵列语音增强的基本原理,总结了各种算法的特点及其所适用的声学环境特性。     

基于麦克风阵列系统的声源定位技术研究现状

麦克风阵列信号处理是数字信号处理领域的一个热点问题,对麦克风阵列接收到的信息量,根据各个阵列之间信息的相关性,可以使用融合处理的方式实现对参数的估计,这种融合不仅可以在时间域处理,也可以在频域处理.麦克风阵列信号处理技术能够在统计意义上得到测量数据,该技术已应用在无线通信,雷达,声纳与工业控制等场合得到了广泛的应用.     

改进的混合下降法在麦克风阵列设计中的应用

在波束形成器设计中,由于麦克风阵列定位优化过程中的非凸性问题,传统的局部搜索技术可能不会产生最优的结果。为了解决这一问题,提出了一种联合遗传算法和梯度方法的混合下降法。通过使用梯度方法在启动点附近迅速找到最优解决方案,同时利用遗传算法避免了局部最小化,从而促进寻找更好的波束形成器设计的最优位置。实验结果表明,与其他几种常用的定位方法相比,使用混合下降方法确定的位置所设计出的波束形成器性能更好。     

QFT application for headphone's active noise cancellation

A quantitative feedback theory design for attenuation of acoustic noise in earphones and headphones is described. Identification including uncertainty is measured for several ears and device location in the ear neighbourhood. Theoretical limitations are discussed. Copyright © 2002 John Wiley & Sons, Ltd.     

On closed-loop adaptive noise cancellation

Given the mean limit ordinary differential equation for the stochastic approximation defining the adaptive algorithm for a closed-loop adaptive noise cancellation, we characterize the limit points. Under appropriate conditions, it is shown that as the dimension of the weight vector increases, the sequence of corresponding limit points converges in the sense of l₂ to the infinite-dimensional optimal weight vector. Also, the limit point of the algorithm is nearly optimal if the dimension of the weight vector is large enough. The gradient of the mean-square error with respect to the weight vector, evaluated at the limit, goes to zero in l₁ and l₂ as the dimension increases, as does the gradient with respect to the coefficients in the transfer function connecting the reference noise signal with the error output. Thus the algorithm is “nearly” a gradient descent algorithm and is error-reducing for large enough dimension. Under broad conditions, iterative averaging can be used to get a nearly optimal rate of convergence     

Adaptive recursive deconvolution and adaptive noise cancellation

In this paper we apply a recursive deconvolution method to active noise cancellation (ANC) in a linear system: the observation of the output of a linear system of relative degree one, read at discrete time instants, is fed to a deconvolution algorithm which identifies the disturbance (with the delay of one step). This information is used in order to reduce the effect of the disturbance itself. Deconvolution being an ill posed problem, a regularization parameter is to be introduced. The choice of the value of the parameter is a delicate issue. We show that, when studying ANC, the discrepancy principle (applied recursively) is a feasible method for the choice of the parameter.     

Blind source separation with convolutive noise cancellation

On-line adaptive learning algorithms for cancellation of additive, convolutive noise from linear mixtures of sources with a simultaneous blind source separation are developed. Associated neural network architectures are proposed. A simple convolutive noise model is assumed, i.e. the unknown additive noise in each channel is a (FIR) filtering version of environmental noise, where some convolutive reference noise is measurable. Two approaches are considered: in the first, the noise is cancelled from the linear mixture of source signals as pre-processing, after that the source signals are separated; in the second, both source separation and additive noise cancellation are performed simultaneously. Both steps consist of adaptive learning processes. By computer simulation experiments, it was found that the first approach is applicable for a large amount of noise, whereas in the second approach, a considerable increase of the convergence speed of the separation process can be achieved. Performance and validity of the proposed approaches are demonstrated by extensive computer simulations.Nomenclature Symbol Meaning - ₄ normalised kurtosis of a signal - (t), t learning rates - m number of sources - n number of sensors - N,M order of the FIR filters - s(t) m-dimensional vector of (unknown) source signals - x(t) n-dimensional vector of mixed signals (sensors) - y(t) n-dimensional vector of separated output signals (estimated sources) - v _R(t) (unknown) primary environment noise signal - n _R(t) secondary reference noise signal - n(t) n-dimensional vector of additive noise signals - f(·),g(·) activation functions in separation rule - f _R(·) activation function in noise cancellation rule - A=[a _ij]_m×n (unknown) mixing matrix - B=[b _ij]_n×N additive noise generation matrix - H(t)=[h _ij]_n×M noise cancellation matrix - W(t)=[w _ij]_n×n global de-mixing matrix     

改进自适应对消算法在工业噪声处理中的应用

分析了工业环境噪声的特点,将自适应噪声对消算法应用到工业噪声的处理当中.在传统最小均方(LMS)算法及基于Lorentzian函数的变步长LMS算法的基础上进一步进行约束稳定性条件处理,提出了一种约束稳定性变步长LMS算法,并在Matlab平台上进行了仿真验证.结果表明:算法具有更快的收敛速度以及更小的稳态误差,并且能有效地降低梯度噪声对算法性能的影响.     

Time-shared channel identification for adaptive noise cancellation in breath sound extraction

Noise artifacts are one of the key obstacles in applying continuous monitoring and computer-assisted analysis of lung sounds. Traditional adaptive noise cancellation (ANC) methodologies work reasonably well when signal and noise are stationary and independent. Clinical lung sound auscultation encounters an acoustic environment in which breath sounds are not stationary and often correlate with noise. Consequently, capability of ANC becomes significantly compromised. This paper introduces a new methodology for extracting authentic lung sounds from noise-corrupted measurements. Unlike traditional noise cancellation methods that rely on either frequency band separation or signal/noise independence to achieve noise reduction, this methodology combines the traditional noise canceling methods with the unique feature of time-sprit stages in breathing sounds. By employing a multi-sensor system, the method first employs a high-pass fdter to elinfinate the off-band noise, and then performs time-shared bfind identification and noise cancellation with recursion from breathing cycle to cycle. Since no frequency separation or signal/noise independence is required, this method potentially has a robust and reliable capability of noise reduction, complementing the traditional methods.     

Adaptive wavelet shrinkage for noise robust speaker recognition

Speaker recognition faces many practical difficulties, among which signal inconsistency due to environmental and acquisition channel factors is most challenging. The noise imposed to the voice signal varies greatly and a priori noise model is usually unavailable. In this article, we propose a robust speaker recognition method that employs a novel adaptive wavelet shrinkage method for noise suppression. In our method, wavelet subband coefficient thresholds are automatically computed, which are proportional to the noise contamination. In the application of wavelet shrinkage for noise removal, a dual-threshold strategy is developed to suppress noise, preserve signal coefficients and minimize the introduction of artifacts. The recognition is achieved using modification of Mel-frequency cepstral coefficient of overlapped voice signal segments. The efficacy of our method is evaluated with voice signals from two public available speech signal databases and is compared with state-of-the-art methods. It is demonstrated that our proposed method exhibits great robustness in various noise conditions. The improvement is significant especially when noise dominates the underlying speech.     

Sparse imputation for large vocabulary noise robust ASR

An effective way to increase noise robustness in automatic speech recognition is to label the noisy speech features as either reliable or unreliable (‘missing’), and replace (‘impute’) the missing ones by clean speech estimates. Conventional imputation techniques employ parametric models and impute the missing features on a frame-by-frame basis. At low SNRs, frame-based imputation techniques fail because many time frames contain few, if any, reliable features. In previous work, we introduced an exemplar-based method, dubbed sparse imputation, which can impute missing features using reliable features from neighbouring frames. We achieved substantial gains in performance at low SNRs for a connected digit recognition task. In this work, we investigate whether the exemplar-based approach can be generalised to a large vocabulary task.Experiments on artificially corrupted speech show that sparse imputation substantially outperforms a conventional imputation technique when the ideal ‘oracle’ reliability of features is used. With error-prone estimates of feature reliability, sparse imputation performance is comparable to our baseline imputation technique in the cleanest conditions, and substantially better at lower SNRs. With noisy speech recorded in realistic noise conditions, sparse imputation performs slightly worse than our baseline imputation technique in the cleanest conditions, but substantially better in the noisier conditions.     

Noise cancellation using selectable adaptive algorithm for speech in variable noise environment

Some of the teething problems associated in the use of two-sensor noise cancellation systems are the nature of the noise signals—a problem that imposes the use of highly complex algorithms in reducing the noise. The usage of such methods can be impractical for many real time applications, where speed of convergence and processing time are critical. At the same time, the existing approaches are based on using a single, often complex adaptive filter to minimize noise, which has been determined to be inadequate and ineffective. In this paper, a new mechanism is proposed to reduce background noise from speech communications. The procedure is based on a two-sensor adaptive noise canceller that is capable of assigning an appropriate filter adapting to properties of the noise. The criterion to achieve this is based on measuring the eigenvalue spread based on the autocorrelation of the input noise. The proposed noise canceller (INC) applies an adaptive algorithm according to the characteristics of the input signal. Various experiments based on this technique using real-world signals are conducted to gauge the effectiveness of the approach. Initial results illustrated the system capabilities in executing noise cancellation under different types of environmental noise. The results based on the INC technique indicate fast convergence rates; improvements up to 30 dB in signal-to-noise ratio and at the same time shows 65% reduction of computational power compared to conventional method.     

Adaptive noise cancellation using enhanced dynamic fuzzy neural networks

In this paper, a novel adaptive noise cancellation algorithm using enhanced dynamic fuzzy neural networks (EDFNNs) is described. In the proposed algorithm, termed EDFNN learning algorithm, the number of radial basis function (RBF) neurons (fuzzy rules) and input-output space clustering is adaptively determined. Furthermore, the structure of the system and the parameters of the corresponding RBF units are trained online automatically and relatively rapid adaptation is attained. By virtue of the self-organizing mapping (SOM) and the recursive least square error (RLSE) estimator techniques, the proposed algorithm is suitable for real-time applications. Results of simulation studies using different noise sources and noise passage dynamics show that superior performance can be achieved.