基于子带卡尔曼滤波的语音增强方法

与基于短时谱的语音增强方法相比,卡尔曼滤波的语音增强方法是基于语音生成模型的增强方法,这种基于模型的递推计算,导致卡尔曼滤波时的计算量很大.为了减少卡尔曼滤波的计算量,本文给出一种基于子带卡尔曼滤波的语音增强方法.先将带噪语音分解成子带信号,并通过子带频域谱减后估计低阶AR模犁参数,然后利用卡尔曼滤波对子带信号进行滤波,最后由滤波后的子带信号重构全带语音信号,实现语音增强.实验表明该方法在提高语音质量的同时,通过子带分解降低了卡尔曼滤波的模型阶数,明显减少了语音增强系统的计算量,更容易实时实现.     

有色噪声下基于Unscented粒子滤波的语音增强方法

针对含有色噪声的语音,提出了一种基于Unscented粒子滤波的单通道语音增强方法.采用时变自回归模型(TVAR)对干净语音建模,通过Unscented粒子滤波器估计AR模型的参数并滤除有色噪声.与大多数常用的粒子滤波选择的建议分布不同,Unscented粒子滤波器采用Unscented卡尔曼滤波器生成粒子滤波的建议分布.由于在粒子的更新过程中考虑了最近的观测值,Unscented粒子滤波器能够在粒子数少于传统粒子滤波算法所需粒子数目的基础上改善估计的性能.仿真实验结果表明,在有色噪声背景下该算法具有良好的语音增强效果.     

结合时变滤波和时频掩码的语音增强方法

本文提出了一种结合鲁棒时变滤波和时频掩码的语音增强方法。首先在带噪语音的时频域中,结合图像处理方法估计出初始瞬时频率信息。然后基于该瞬时频率信息,利用鲁棒时变滤波算法构建降噪后的语音信号。最后根据重构语音的时频特征预测时频掩码。该掩码在带噪语音的时频域中能够有效地保留语音成分且抑制噪声成分,从而达到语音增强的目的。实验结果表明,在几种常见背景噪声环境下,所提语音增强算法在抑制背景噪声干扰、提升语音整体质量方面表现良好,尤其是在低信噪比环境下具有明显的优势。      

基于临界频带的子带滤波语音增强

根据人耳的听觉感知特性,提出了一种基于子带滤波的优化语音增强方法。基于临界频带设计滤波器将输入信号分成若干子带,依据估计出的每个子带的短时信噪比来对相应子带的时域信号逐帧进行独立的自适应处理后再合成。语音增强性能评估结果表明,有效地去除背景噪声的同时还抑制了音乐噪声,减少了语音的听觉失真,提高了增强语音的可懂度。     

基于蒙特卡洛滤波平滑的语音增强算法

本文在分析统计信号贝叶斯模型和语音信号的时变自回归(TVAR)模型的基础上,利用蒙特卡洛滤波及平滑方法,对语音信号的TVAR模型参数进行了估计,提出了一种有效的针对非平稳加性噪声影响下的语音增强算法.该算法可以很好的跟踪非平稳信号,同时引入对反射系数的判断,保证了跟踪的稳定性.实验表明,本文方法能很好的抑制背景噪声,提高信噪比,改善语音信号的听觉质量.     

基于HFM和感知后滤波器的语音增强

研究了只能获得带噪信号的情况下的语音增强问题。将语音信号看作由高斯噪声激励的自回归(AR)过程,观测噪声为加性高斯白噪声,把信号转化为状态空间模型。首先用隐马尔可夫模型(HMM)估计AR参数和噪声的方差作为卡尔曼滤波器初值,估计信号作为参数估计的中间值给出,然后将估计信号通过一个感知滤波器平滑以消除残余噪声。仿真结果表明该算法有良好的性能。     

基于改进粒子滤波的电力线载波通信语音增强

针对噪声和混响环境下的电力线载波机语音通信系统聆听上的困难,研究了为实现多路电力线载波机复接器所采用的粒子滤波的语音增强算法,提出了一种基于粒子群优化的改进粒子滤波算法,它将语音增强问题转换为从带噪语音中对纯净语音的估计过程,引入粒子群优化的方法来产生建议分布,使降噪结果更接近纯净语音,从而得到更好的语音增强效果。     

一种基于子带GSC的语音增强算法

提出一种基于GSC的语音增强算法,该算法应用了DFT调制子带滤波器组将语音信号分解到子带进行自适应滤波,从而获得更好的增强效果以及更低的运量复杂度.同时,将范数约束自适应滤波(NCAF)算法应用于自适应噪声对消器(ANC)以降低语音的失真度.为了进一步去除增强后语音中的残留噪声,算法使用改进的Wiener后置滤波器.仿真结果表明,相对于基于全带GSC的麦克风阵列语音增强算法以及传统Wiener后置滤波算法,采用本文所用算法具有更高的输出分段信噪比.     

一种新的含噪混沌信号降噪算法

该文针对低信噪比、非高斯加性噪声和混沌动力学系统参数未知的含噪混沌信号降噪问题,提出了一种基于粒子滤波(Particle Filtering, PF)的降噪新算法。该算法将混沌信号和动力学系统中的未知参数作为一个多维状态矢量,利用PF方法递推计算多维状态矢量的联合后验概率分布,进而实现了对混沌信号的最优估计。对于混沌信号轨道分离过快所导致的退化问题,提出了有效的解决方法,并利用核平滑和自回归(Auto-Regression, AR)模型建模的方法分别实现了非时变以及时变参数的递推估计。仿真实验的结果表明,与现有的降噪方法相比,该文提出的新算法能够更加有效地抑制含噪混沌信号中的加性噪声。     

单载波频域均衡系统信道估计的粒子滤波方法

粒子滤波（Particle Filter, PF）是一种有效的参数估计方法。通过对单载波频域均衡（Single Carrier Frequency Domain Equalization, SC-FDE）系统数学模型和粒子滤波原理的分析,将时变信道建模成一阶AR过程,尝试把粒子滤波方法运用到单载波频域均衡系统基于UW的信道估计中去,并给出了算法详细步骤。然后,分别针对三种不同时变程度的信道进行了仿真,并在这三种信道下,分别与LS估计作了误码性能比较。结果表明,在时变条件下,基于粒子滤波的信道估计方法较之线性LS估计能获得良好的误码性能增益,且信道变化越缓慢,这种增益越明显。      

A Novel Method for Partially Adaptive Broadband Beamforming

In this paper, a novel subband-selective generalized sidelobe canceller (GSC) for partially adaptive broadband beamforming is proposed. The blocking matrix of the GSC is constructed such that its columns constitute a series of bandpass filters, which select signals with specific angles of arrival and frequencies. This results in bandlimited spectra of the blocking matrix outputs, which is further exploited by a subband decomposition prior to running independent unconstrained adaptive filters in each non-redundant subband. We discuss the design of both the blocking matrix using a genetic algorithm for an efficient sum-of-power-of-two coefficient format and the filter bank for the subsequent subband decomposition. By these steps, the computational complexity of our subband-selective GSC is greatly reduced compared to other adaptive GSC schemes, while performance is comparable or even enhanced due to subband decorrelation, as simulations indicate.     

Novel efficient approach for the design of equiripple quadraturemirror filters

Quadrature mirror filters have been used extensively in subband coding of speech signals. The authors introduce a novel efficient approach for the design of equiripple quadrature mirror filters. The new approach is more efficient than the previously proposed design method in terms of computer time and memory requirement     

Subband adaptive filtering with real-valued subband signals foracoustic echo cancellation

Adaptive filtering in subbands was originally proposed to overcome the limitations of conventional least-mean-square (LMS) algorithms. In general, subband adaptive filters offer computational savings, as well as faster convergence over the conventional LMS algorithm. However, improvements to current subband adaptive filters could be further enhanced by a more elegant choice of their design/structure. Classical subband adaptive filters employ DFT-based analysis and synthesis filter banks which results in subband signals that are complex-valued. The authors modify the structure of subband adaptive filters by using single-sideband (SSB) modulated analysis and synthesis filter banks, which result in subband signals that are real-valued. This simplifies the realisation of subband adaptive filters     

一种改进的奇异值分解语音增强方法

该文将多麦克语音增强方法用于单麦克语音增强,给出了一种改进的奇异值分解语音增强方法。该方法首先对输入矩阵进行雅克比奇异值分解,用得到的奇异值矢量构造语音增强滤波器;然后用输入矩阵与滤波器权矢量相乘来构造各路信号;最后采用麦克风阵列波束形成的方法,得到增强后的语音信号。仿真结果表明,该方法能有效地去除加性噪声,并且改善了语音质量。     

Speech Enhancement Using Perceptual Wavelet Packet Decomposition and Teager Energy Operator

It has been shown in the literature that the perceptual wavelet packet decomposition (PWPD) and the Teager energy operator (TEO) are useful for various speech processing systems and speech enhancement applications, respectively. By the use of the PWPD and the TEO, this paper presents an improved wavelet-based speech enhancement method. The main advantage of the proposed method is that the over thresholding of speech segments which is usually occurred in conventional wavelet-based speech enhancement schemes can be avoided. As a consequence, the enhanced speech quality of the proposed method can be increased substantially from those of conventional approaches. In addition, the proposed method does not require a complicated estimation of the noise level or any knowledge of the SNR. Using speech signals corrupted by additive and real noises, experimental results demonstrate that the speech enhancement method presented in this paper is capable of outperforming conventional noise cancellation schemes.     

A subband-selective broadband GSC with cosine-modulated blocking matrix

A novel subband-selective generalized sidelobe canceller (GSC) for partially adaptive broadband beamforming is proposed. The columns of the blocking matrix are derived from a prototype vector by cosine modulation, and the broadside constraint is incorporated by imposing zeros on the prototype vector appropriately. These columns constitute a series of bandpass filters, which select signals with specific directions of arrival and frequencies. This results in a high-pass-type bandlimited spectra of the blocking matrix outputs, which is further exploited by subband decomposition and suitably discarding the low-pass subbands prior to running independent unconstrained adaptive filters in each nonredundant subband. By these steps, the computational complexity of a GSC implementation is greatly reduced compared to fully adaptive GSC schemes, while performance is comparable or even enhanced due to subband decorrelation in both spatial and temporal domains.     

Linear prediction of subband signals

The performance of linear prediction of fullband and subband signals is described in terms of the respective prediction gain. The subband prediction gain is characterized in terms of the fullband signal power spectral density and the frequency response of the subband filters. For Gaussian fullband signals, the asymptotic subband prediction gain can never be larger than the asymptotic fullband prediction gain. Simulation results compare fixed and adaptive fullband and subband prediction gains for Gaussian sources and speech. For speech, the subband prediction gain can exceed the fullband prediction gain     

A Low-Complexity Delayless Selective Subband Adaptive Filtering Algorithm

Adaptive filters of significant order, requiring high computational complexity, are necessary in many applications such as acoustic echo cancellation and wideband active noise control. Successful approaches to lessen the computational complexity of such filters are subband methods, and partial updating schemes where only a part of the filter is updated at each instant. To avoid the time delay introduced by the subband-splitting, delayless structures which reconstructs a fullband filter, producing delayless output, from the adaptive subband filters have been proposed. This paper proposes a delayless subband adaptive filter partial updating scheme, where the general idea is to only update the most misadjusted subband filter(s). Analysis in terms of mean square deviation is presented and shows that the fullband filter convergence speed is significantly increased, even for flat spectrum signals, as compared to traditional periodic subband filter update with the same computational complexity. Echo cancellation simulations with an artificial system to verify the analysis, using both flat spectrum signals and speech, is also presented, as well as offline calculations using signals from a real system.