拉普拉斯分布下的MMSE谱减语音增强算法

针对基于高斯分布的谱减语音增强算法,增强语音出现噪声残留和语音失真的问题,提出了基于拉普拉斯分布的最小均方误差（MMSE）谱减算法。首先,对原始带噪语音信号进行分帧、加窗处理,并对处理后每帧的信号进行傅里叶变换,得到短时语音的离散傅里叶变换（DFT）系数;然后,通过计算每一帧的对数谱能量及谱平坦度,进行噪声帧检测,更新噪声估计;其次,基于语音DFT系数服从拉普拉斯分布的假设,在最小均方误差准则下,求解最佳谱减系数,使用该系数进行谱减,得到增强信号谱;最后,对增强信号谱进行傅里叶逆变换、组帧,得到增强语音。实验结果表明,使用所提算法增强的语音信噪比（SNR）平均提高了4.3 dB,与过减法相比,有2 dB的提升;在语音质量感知评估（PESQ）得分方面,与过减法相比,所提算法平均得分有10%的提高。该算法有更好的噪声抑制能力和较小的语音失真,在SNR和PESQ评价标准上有较大提升。     

基于感知掩蔽深度神经网络的单通道语音增强方法

本文将心理声学掩蔽特性应用于基于深度神经网络（Deep neural network,DNN）的单通道语音增强任务中,提出了一种具有感知掩蔽特性的DNN结构.首先,提出的DNN对带噪语音幅度谱特征进行训练并分别得到纯净语音和噪声的幅度谱估计.其次,利用估计的纯净语音幅度谱计算噪声掩蔽阈值.然后,将噪声掩蔽阈值和估计的噪声幅度谱联合计算得到一个感知增益函数.最后,利用感知增益函数从带噪语音幅度谱中估计出增强语音幅度谱.在TIMIT数据库上,对不同信噪比下的20种噪声进行的仿真实验表明,无论噪声类型是否在语音的训练集中出现,所提出的感知掩蔽DNN都能够在有效去除噪声的同时保持较小的语音失真,增强效果明显优于常见的DNN增强方法以及NMF（Nonnegative matrix factorization）增强方法.     

Speech Enhancement Based on Generalized Minimum Mean Square Error Estimators and Masking Properties of the Auditory System

In this paper, the family of conditional minimum mean square error (MMSE) spectral estimators is studied which take on the form$(E(X_p^alpha/vert X_p+D_pvert))^1/alpha$, where$X_p$is the clean speech spectrum, and$D_p$is the noise spectrum, resulting in a Generalized MMSE estimator (GMMSE). The degree of noise suppression versus musical tone artifacts of these estimators is studied. The tradeoffs in selection of$(alpha)$, across noise spectral structure and signal-to-noise ratio (SNR) level, are also considered. Members of this family of estimators include the Ephraim–Malah (EM) amplitude estimator and, for high SNRs, the Wiener Filter. It is shown that the colorless residual noise observed in the EM estimator is a characteristic of this general family of estimators. An application of these estimators in an auditory enhancement scheme using the masking threshold of the human auditory system is formulated, resulting in the GMMSE-auditory masking threshold (AMT) enhancement method. Finally, a detailed evaluation of the proposed algorithms is performed over the phonetically balanced TIMIT database and the National Gallery of the Spoken Word (NGSW) audio archive using subjective and objective speech quality measures. Results show that the proposed GMMSE-AMT outperforms MMSE and log-MMSE enhancement methods using a detailed phoneme-based objective quality analysis.     

Adaptive Time Segmentation for Improved Speech Enhancement

Single-channel enhancement algorithms are widely used to overcome the degradation of noisy speech signals. Speech enhancement gain functions are typically computed from two quantities, namely, an estimate of the noise power spectrum and of the noisy speech power spectrum. The variance of these power spectral estimates degrades the quality of the enhanced signal and smoothing techniques are, therefore, often used to decrease the variance. In this paper, we present a method to determine the noisy speech power spectrum based on an adaptive time segmentation. More specifically, the proposed algorithm determines for each noisy frame which of the surrounding frames should contribute to the corresponding noisy power spectral estimate. Further, we demonstrate the potential of our adaptive segmentation in both maximum likelihood and decision direction-based speech enhancement methods by making a better estimate of the a priori signal-to-noise ratio (SNR)$xi$. Objective and subjective experiments show that an adaptive time segmentation leads to significant performance improvements in comparison to the conventionally used fixed segmentations, particularly in transitional regions, where we observe local SNR improvements in the order of 5 dB.     

Improved Signal-to-Noise Ratio Estimation for Speech Enhancement

This paper addresses the problem of single-microphone speech enhancement in noisy environments. State-of-the-art short-time noise reduction techniques are most often expressed as a spectral gain depending on the signal-to-noise ratio (SNR). The well-known decision-directed (DD) approach drastically limits the level of musical noise, but the estimated a priori SNR is biased since it depends on the speech spectrum estimation in the previous frame. Therefore, the gain function matches the previous frame rather than the current one which degrades the noise reduction performance. The consequence of this bias is an annoying reverberation effect. We propose a method called two-step noise reduction (TSNR) technique which solves this problem while maintaining the benefits of the decision-directed approach. The estimation of the a priori SNR is refined by a second step to remove the bias of the DD approach, thus removing the reverberation effect. However, classic short-time noise reduction techniques, including TSNR, introduce harmonic distortion in enhanced speech because of the unreliability of estimators for small signal-to-noise ratios. This is mainly due to the difficult task of noise power spectrum density (PSD) estimation in single-microphone schemes. To overcome this problem, we propose a method called harmonic regeneration noise reduction (HRNR). A nonlinearity is used to regenerate the degraded harmonics of the distorted signal in an efficient way. The resulting artificial signal is produced in order to refine the a priori SNR used to compute a spectral gain able to preserve the speech harmonics. These methods are analyzed and objective and formal subjective test results between HRNR and TSNR techniques are provided. A significant improvement is brought by HRNR compared to TSNR thanks to the preservation of harmonics.     

基于听觉掩蔽效应的MMSE语音增强算法

针对MMSE语音增强算法低信噪比时产生较大的语音畸变的缺点,提出了一种结合人耳听觉掩蔽效应的MMSE语音增强算法。该算法利用掩蔽阈值来调整MMSE算法中的增益值,使得增强后的语音信号残留噪声和语音畸变较小。通过计算机仿真对增强前后语音信号的信噪比分析以及主观试听表明：改进的MMSE语音增强算法不仅提高了语音信号的信噪比,而且减少了语音畸变,提高了语音的可懂度。     

基于噪声估计的二值掩蔽语音增强算法

针对现有的助听器语音增强算法在非平稳噪声环境下,残留大量背景噪声的同时还引入了“音乐噪声”,致使增强语音可懂度和信噪比不理想等问题。提出了一种基于噪声估计的二值掩蔽语音增强算法,该算法利用人耳听觉感知理论,结合人耳的听觉特性和耳蜗的工作机理。采用最小值控制递归平均（Minima-Controlled Recursive Averaging,MCRA）算法获得估计噪声和初步增强语音;将估计噪声和初步增强语音分别通过可以模拟人工耳蜗模型的gammatone滤波器组进行滤波处理,得到各自的时频表示形式;利用人耳的听觉掩蔽特性,计算含噪语音在时频域的二值掩蔽;利用二值掩蔽得到增强语音。实验结果表明：该算法很大程度上去除了谱减法引入的“音乐噪声”,与基于MCRA谱减法相比,增强语音的语言可懂度指数（Speech Intelligibility Index,SII）、主观语音质量评估（Perceptual Evaluation of Speech Quality,PESQ）和信噪比（Signal to Noise Ratio,SNR）都得到了提高。     

Text-Independent Speaker Verification for Real Fast-Varying Noisy Environments

Investigating Speaker Verification in real-world noisy environments, a novel feature extraction process suitable for suppression of time-varying noise is compared with a fine-tuned spectral subtraction method. The proposed feature extraction process is based on approximating the clean speech and the noise spectral magnitude with a mixture of Gaussian probability density functions (pdfs) by using the Expectation-Maximization algorithm (EM). Subsequently, the Bayesian inference framework is applied to the degraded spectral coefficients, and by employing Minimum Mean Square Error Estimation (MMSE), a closed form solution for the spectral magnitude estimation task is derived. The estimated spectral magnitude finally is incorporated into the Mel-Frequency Cepstral Coefficients (MFCCs) front-end of a baseline text-independent speaker verification system, based on Probabilistic Neural Networks, which participated successfully in the 2002 NIST (National Institute of Standards and Technology of USA) Speaker Recognition Evaluation. A comparative study of the proposed technique for real-world noise types demonstrates a significant performance gain compared to the baseline speech features and to the spectral subtraction enhancement method. Improvements of the absolute speaker verification performance with more than 27% for 0 dB signal-to-noise ratio (SNR), compared to the MFCCs, and with more than 13% for –5 dB SNR, compared to the spectral subtraction version, were obtained in the case of a passing-by aircraft scenario.     

Minimum Mean-Square Error Estimation of Discrete Fourier Coefficients With Generalized Gamma Priors

This paper considers techniques for single-channel speech enhancement based on the discrete Fourier transform (DFT). Specifically, we derive minimum mean-square error (MMSE) estimators of speech DFT coefficient magnitudes as well as of complex-valued DFT coefficients based on two classes of generalized gamma distributions, under an additive Gaussian noise assumption. The resulting generalized DFT magnitude estimator has as a special case the existing scheme based on a Rayleigh speech prior, while the complex DFT estimators generalize existing schemes based on Gaussian, Laplacian, and Gamma speech priors. Extensive simulation experiments with speech signals degraded by various additive noise sources verify that significant improvements are possible with the more recent estimators based on super-Gaussian priors. The increase in perceptual evaluation of speech quality (PESQ) over the noisy signals is about 0.5 points for street noise and about 1 point for white noise, nearly independent of input signal-to-noise ratio (SNR). The assumptions made for deriving the complex DFT estimators are less accurate than those for the magnitude estimators, leading to a higher maximum achievable speech quality with the magnitude estimators.     

Robust feature extraction from spectrum estimated using bispectrum for speaker recognition

Extraction of robust features from noisy speech signals is one of the challenging problems in speaker recognition. As bispectrum and all higher order spectra for Gaussian process are identically zero, it removes the additive white Gaussian noise while preserving the magnitude and phase information of original signal. The spectrum of original signal can be recovered from its noisy version using this property. Robust Mel Frequency Cepstral Coefficients (MFCC) are extracted from the estimated spectral magnitude (denoted as Bispectral-MFCC (BMFCC)). The effectiveness of BMFCC has been tested on TIMIT and SGGS databases in noisy environment. The proposed BMFCC features yield 95.30?%, 97.26?% and 94.22?% speaker recognition rate on TIMIT, SGGS and SGGS2 databases, respectively for 20 dB SNR whereas these values for 0 dB SNR are 45.84?%, 50.79?% and 44.98?%. The experimental results show the superiority of the proposed technique compared to conventional methods for all databases.     

Single channel noise reduction system in low SNR

We propose a two stage noise reduction system for reducing background noise using single-microphone recordings in very low signal-to-noise ratio (SNR) based on Wiener filtering and ideal binary masking. The proposed system contains two stages. In first stage, the Wiener filtering with improved a priori SNR is applied to noisy speech for background noise reduction. In second stage, the ideal binary mask is estimated at every time–frequency channel by using pre-processed first stage speech and comparing the time–frequency channels against a pre-selected threshold T to reduce the residual noise. The time–frequency channels satisfying the threshold are preserved whereas all other time–frequency channels are attenuated. The results revealed substantial improvements in speech intelligibility and quality over that accomplished with the traditional noise reduction algorithms and unprocessed speech.     

Audio enhancement using local SNR-based sparse binary mask estimation and spectral imputation

This paper proposes a method for enhancing speech and/or audio quality under noisy conditions. The proposed method first estimates the local signal-to-noise ratio (SNR) of the noisy input signal via sparse non-negative matrix factorization (SNMF). Next, a sparse binary mask (SBM) is proposed that separates the audio signal from the noise by measuring the sparsity of the pool of local SNRs from the adjacent frequency bands of the current and several previous frames. However, some spectral gaps remain across frequency bands after applying the binary masks, which distorts the separated audio signal due to spectral discontinuity. Thus, a spectral imputation technique is used to fill the empty spectrum of the frequency band where it is removed by the SBM. Spectral imputation is conducted by online learning NMF with the spectra of the neighboring non-overlapped frequency bands and their local sparsity. The effectiveness of the proposed enhancement method is demonstrated on two different tasks use speech and musical content, respectively. Consequently, objective measurements and subjective listening tests show that the proposed method outperforms conventional speech and audio enhancement methods, such as SNMF-based alternatives and deep recurrent neural networks for speech enhancement, block thresholding, and a commercially available software tool for audio enhancement.     

A gain factor adapted by masking property and SNR variation for speech enhancement in colored-noise corruptions

A gain factor adapted by both the intra-frame masking properties of the human auditory system and the inter-frame SNR variation is proposed to enhance a speech signal corrupted by additive noise. In this article we employ an averaging factor, varying with time–frequency, to improve the estimate of the a priori SNR. In turn, this SNR estimate is utilized to adapt a gain factor for speech enhancement. This gain factor reduces the spectral variation over successive frames, so the effect of musical residual noise is mitigated. In addition, the simultaneous masking property of the human ears is also employed to adapt the gain factor. Imperceptive residual noise with energy below the noise masking threshold is retained, resulting in a reduction of speech distortion. Experimental results show that the proposed scheme can efficiently reduce the effect of musical residual noise.     

Speech Enhancement, Gain, and Noise Spectrum Adaptation Using Approximate Bayesian Estimation

This paper presents a new approximate Bayesian estimator for enhancing a noisy speech signal. The speech model is assumed to be a Gaussian mixture model (GMM) in the log-spectral domain. This is in contrast to most current models in frequency domain. Exact signal estimation is a computationally intractable problem. We derive three approximations to enhance the efficiency of signal estimation. The Gaussian approximation transforms the log-spectral domain GMM into the frequency domain using minimal Kullback-Leiber (KL)-divergency criterion. The frequency domain Laplace method computes the maximum a posteriori (MAP) estimator for the spectral amplitude. Correspondingly, the log-spectral domain Laplace method computes the MAP estimator for the log-spectral amplitude. Further, the gain and noise spectrum adaptation are implemented using the expectation-maximization (EM) algorithm within the GMM under Gaussian approximation. The proposed algorithms are evaluated by applying them to enhance the speeches corrupted by the speech-shaped noise (SSN). The experimental results demonstrate that the proposed algorithms offer improved signal-to-noise ratio, lower word recognition error rate, and less spectral distortion.     

基于语音存在概率和听觉掩蔽特性的语音增强算法

低信噪比下,谱减语音增强法中一直存在的去噪度、残留的音乐噪声和语音畸变度三者间均衡这一关键问题显得尤为突出。为降低噪声对语音通信的干扰,提出了一种适于低信噪比下的语音增强算法。在传统的谱减法基础上,根据噪声的听觉掩蔽阈值自适应调整减参数,利用语音存在概率,对语音、噪声信号估计,避免低信噪比下端点检测(VAD)的不准确,有更强的鲁棒性。对算法进行了客观和主观测试,结果表明：相对于传统的谱减法,在几乎不损伤语音清晰度的前提下该算法能更好地抑制残留噪声和背景噪声,特别是对低信噪比和非平稳噪声干扰的语音信号,效果更加明显。     

一种新的基于信息熵的带噪语音端点检测方法

在自动语音识别和变速率语音编码技术中,语音端点检测是前端处理的一个重要环节.而在实际的噪声环境下,一些传统的端点检测方法已不适用.该文提出了一种新的基于信息熵的语音端点检测方法,该方法通过对语音信号的短时功率谱进行谱分析,由此构造熵函数作为端点检测的特征参数.实验结果表明,该方法在噪声环境下性能优于传统的基于能量的端点检测方法.而且相对于基于频谱谱熵的算法,在低信噪比（SNR〈0dB）情况下,该文方法有更好的鲁棒性,可使平均检测精确度进一步提高约5%.     

Target Speech Detection and Separation for Communication with Humanoid Robots in Noisy Home Environments

People usually talk face to face when they communicate with their partner. Therefore, in robot audition, the recognition of the front talker is critical for smooth interactions. This paper presents an enhanced speech detection method for a humanoid robot that can separate and recognize speech signals originating from the front even in noisy home environments. The robot audition system consists of a new type of voice activity detection (VAD) based on the complex spectrum circle centroid (CSCC) method and a maximum signal-to-noise ratio (SNR) beamformer. This VAD based on CSCC can classify speech signals that are retrieved at the frontal region of two microphones embedded on the robot. The system works in real-time without needing training filter coefficients given in advance even in a noisy environment (SNR > 0 dB). It can cope with speech noise generated from televisions and audio devices that does not originate from the center. Experiments using a humanoid robot, SIG2, with two microphones showed that our system enhanced extracted target speech signals more than 12 dB (SNR) and the success rate of automatic speech recognition for Japanese words was increased by about 17 points.     

联合优化深度神经网络和约束维纳滤波的单通道语音增强方法

深度神经网络（Deep neural networks,DNNs）依靠其良好的特征提取能力,在语音增强任务中得到了广泛应用。为进一步提高深度神经网络的语音增强效果,提出一种将深度神经网络和约束维纳滤波联合训练优化的新型网络结构。该网络首先对带噪语音幅度谱进行训练并分别得到纯净语音和噪声的幅度谱估计,然后利用语音和噪声的幅度谱估计计算得到一个约束维纳增益函数,最后利用约束维纳增益函数从带噪语音幅度谱中估计出增强语音幅度谱作为网络的训练输出。对不同信噪比下的20种噪声进行的仿真实验表明,无论噪声类型是否在网络的训练集中出现,本文方法都能够在有效去除噪声的同时保持较小的语音失真,增强效果明显优于DNN及NMF增强方法。     

基于语音增强失真补偿的抗噪声语音识别技术

本文提出了一种基于语音增强失真补偿的抗噪声语音识别算法。在前端,语音增强有效地抑制背景噪声;语音增强带来的频谱失真和剩余噪声是对语音识别不利的因素,其影响将通过识别阶段的并行模型合并或特征提取阶段的倒谱均值归一化得到补偿。实验结果表明,此算法能够在非常宽的信噪比范围内显著的提高语音识别系统在噪声环境下的识别精度,在低信噪比情况下的效果尤其明显,如对-5dB的白噪声,相对于基线识别器,该算法可使误识率下降67.4%。     

A novel fast nonstationary noise tracking approach based on MMSE spectral power estimator

Estimating the noise power spectral density (PSD) from the corrupted speech signal is an essential component for speech enhancement algorithms. In this paper, a novel noise PSD estimation algorithm based on minimum mean-square error (MMSE) is proposed. The noise PSD estimate is obtained by recursively smoothing the MMSE estimation of the current noise spectral power. For the noise spectral power estimation, a spectral weighting function is derived, which depends on the a priori signal-to-noise ratio (SNR). Since the speech spectral power is highly important for the a priori SNR estimate, this paper proposes an MMSE spectral power estimator incorporating speech presence uncertainty (SPU) for speech spectral power estimate to improve the a priori SNR estimate. Moreover, a bias correction factor is derived for speech spectral power estimation bias. Then, the estimated speech spectral power is used in “decision-directed” (DD) estimator of the a priori SNR to achieve fast noise tracking. Compared to three state-of-the-art approaches, i.e., minimum statistics (MS), MMSE-based approach, and speech presence probability (SPP)-based approach, it is clear from experimental results that the proposed algorithm exhibits more excellent noise tracking capability under various nonstationary noise environments and SNR conditions. When employed in a speech enhancement system, improved speech enhancement performances in terms of segmental SNR improvements (SSNR+) and perceptual evaluation of speech quality (PESQ) can be observed.