A fast adaptive Kalman filtering algorithm for speech enhancement

We propose a speech enhancement algorithm that applies a Kalman filter in the modulation domain to the output of a conventional enhancer operating in the time-frequency domain. We show that the prediction residual signal of the spectral amplitude errors at the output of the baseline MMSE enhancer do not follow a Gaussian distribution. Accordingly, the Kalman filter used in our enhancement algorithm combines a colored noise model with a Gaussian mixture model of the residual noise. We evaluate the performance of the speech enhancement algorithm on the core TIMIT test set and demonstrate that it gives consistent performance improvements over the baseline enhancer and over a previously proposed Kalman filter post-processor. Index Terms—speech enhancement, post-processing, Kalman filter, Gaussian mixture model, modulation domain

The family of state space-based speech enhancement algorithms is taken as the central building block for this thesis; such algorithms include for example Kalman Filters running on autoregressive models of speech. The goal of this thesis is to render this family of algorithms ready, that is, upgrade the existing solutions to make them more flexible and more robust to real-world situations where different types of nonstationary, non-Gaussian noise are to be expected. The chosen context is that of sensitive applications such as hearing aids for which naturalness, intelligibility, and noise reduction are equally important. Most of the state space-based speech enhancement algorithms are currently unable to operate robustly in nonstationary colored-noise environments, and those found in the literature usually assume AWGN conditions. Since in addition, they are in general more computationally demanding than other well-established algorithms, it is not surprising that they are often either overlooked or omitted in respected and recent speech enhancement publications, as the introduction in this thesis will underline. Nevertheless, for AWGN conditions they have still been reported to yield high quality and natural-sounding results, making them an appealing choice for the range of applications chosen. To achieve the goal of this thesis, it is first necessary to categorize, implement and test many state-space based algorithms for speech enhancement; as it turns out, most of the existing algorithms revolve around the same principles and ideas. Then, in this work several tools and extensions of state spaced-based algorithms are developed in order to tackle the required handling of real-world noise. In parallel, since one of the main goals is to preserve naturalness and intelligibility, a variety of configurations based on these proposed tools and extensions are concocted and thoroughly tested so as to be able to determine the best solution both in terms of output quality and complexity requirements. As an additional constraint, most of the tools devised are meant to be compatible with any state space-based algorithm, rendering the work in this thesis homogeneous and unifying. To deal with real-world noise, novel fullband and subband solutions that take into account information from existing noise estimation algorithms are proposed. In the fullband domain, new all-pole modelling techniques are devised, and in the subband domain several alternatives to handle the noise are proposed as well. Throughout the thesis, several important accessory contributions are also developed, such as a new category of particle filtering algorithms for nonlinear speech models, low-cost post-processing techniques for further noise reduction, binaural extensions of monaural state-space models, and various improvement ideas specifically targeted to state-space algorithms running on autoregressive speech models. As an example, a possible state-space based algorithm is proposed based on some of the conclusions of the several Chapters of this thesis, which is shown to outperform three well-established state-of-the-art algorithms in adverse conditions. Rather than merely justifying the work in this thesis by a final arm-wrestling contest, these latter results prove that state-space based algorithms deserve a place amongst highly regarded and viable algorithms for speech enhancement.

This paper addresses the speech enhancement problem with adaptive filtering algorithms. We propose a new dual forward blind source separation (FBSS) algorithm based on the use of the recursive least square algorithm to update the cross-filters of the forward structure. This algorithm inherits the good characteristics of the combination between the FBSS and the good properties of the RLS algorithm. In this work, we propose to use the DFRLS algorithm in speech enhancement and acoustic noise reduction application. This proposed algorithm shows good characteristics in comparison with dual forward normalized least mean square (DFNLMS) algorithm is terms of various objective criteria as the segmental signal to noise ratio (SegSNR), the cepstral distance (CD), the system mismatch (SM) and the segmental mean square error (SegMSE).

Most algorithms for speech enhancement in the spectral domain focus on the acquisition of an estimator of the clean speech parameter, such as spectrum or amplitude. Enhanced speech quality with these methods mainly depends upon the accuracy of the estimator. When the signal-to-noise ratio (SNR) becomes lower, e.g., less than 3 dB, the enhanced speech often shows unsatisfactory quality. Here, we propose a new method for speech enhancement in the spectral domain. It is based on our view that speech characteristics are perceived mainly by part of the spectral components which have higher local (or instantaneous) SNRs. These components have a special importance on speech SNR enhancement and speech understanding. Identifying these components in the noisy spectrum with a decision system and constructing enhanced speech just by the same noisy amplitude and phase, the result is similar to those popular algorithms. All that is needed is a local SNR decision for each spectral component. The threshold for the decision is derived and two popular methods for local SNR (LSNR) estimation are suggested. It shows better performance under lower SNR signals. Though the number of these kind of spectral components varies with different signal SNR, any algorithms which can pick up more than 90% of it is sufficient to match any existing methods. Primary considerations and results are shown.

A fast complex-valued adaptive filtering algorithm based on the Euclidean direction set (EDS) method is presented in this paper. A mathematical analysis of the complex-valued EDS (CEDS) algorithm is also performed in this paper. It is shown that the CEDS algorithm has a computational complexity of O ( N ) for each system update, and that the CEDS converges to the true solution. Computer simulations illustrate that this complex algorithm is very efficient in complex adaptive FIR and IIR filtering.

Combined speech enhancement and auditory modelling for robust distributed speech recognition

Automatic separation of speech from noise and segregation of overlapping co-channel speech are two of the most challenging problems in speech processing. In previous work, we have developed algorithms to both enhance noisy speech and segregate overlapping speech streams. Our single-channel speech enhancement and speech segregation algorithms have shown better performance than other reported algorithms for automatic speech recognition. Additionally, objective evaluation scores of perceptual quality have shown a significant improvement following processing by our algorithms. In this study, we focus on subjective evaluation of these algorithms for human listeners. We investigate the intelligibility and quality of speech from our algorithm on normal-hearing listeners, cochlear implant users, and hearing impaired subjects. Our preliminary results indicate a significant improvement in the perceptual quality of the speech signal after being processed by our algorithm and suggest that the proposed algorithms can be used as a pre-processing block within the signal processing in hearing aid devices.

In this paper, we propose a statistical model-based speech enhancement algorithm using an improved minima controlled recursive averaging (IMCRA) noise estimation and a decision-directed (DD) priori SNR estimation. In the training stage, the Gaussian mixture model (GMM) of the Mel-frequency cepstral coefficients (MFCCs) of universal speaker is obtained. In speech enhancement stage, minima tracking process of IMCRA noise estimation is adjusted with the noisy power spectrum of current frame and an adjustment weighting factor. In addition, based on the universal GMM, some significant constant parameters are replaced by frequency-varying parameters, such as the weighting parameter in the DD priori SNR estimation and the adjustment weighting factor in the modified minima tracking process of IMCRA. The performance of proposed speech enhancement is evaluated by objective tests under various stationary and non-stationary noise environments. From experimental results, compared to the conventional approaches, the proposed scheme performs better and is suitable for being used as the pre-processing of speech processing systems.

This paper addresses the problem of speech enhancement and acoustic noise reduction by partial and set-membership adaptive algorithms combined with the symmetric decorrelating adaptive (SAD) algorithm structure. In this paper, we propose two new adaptive algorithms based on set-membership principle that improve the original set-membership algorithm behavior in speech enhancement applications. The first proposed algorithm (called Proposed 1), is based on the combination of the SAD algorithm structure with a smart control that uses decorrelating properties between the output and the mixing signal to control and update the SAD adaptive filters. The second proposed algorithm (called Proposed 2) is a modification of Proposed 1 and based on a new regularization relations of the SAD adaptive filters that use a combination between the variance of the mixing and the output signals of the SAD structures. These two proposed algorithms (Proposed 1 and Proposed 2) aim to improve the convergence speed performance and the output signal-to-noise-ratio of the original SAD algorithm when no smart control of the adaptive filters is used. The proposed algorithms have very interesting properties with non-stationary signal like speech when the SAD algorithm is, used alone, fails. The simulation results that are obtained by the comparison between the proposed algorithms (Proposed 1 and Proposed 2) and the original two-channel set-membership NLMS algorithm have shown the best performances of Proposed 1 and Proposed 2 in terms of the following criteria: systems mismatch, segmental SNR and segmental men square error.

Speech signal often gets degraded by various types of noise at different stages of speech signal operations. One major type of noise is the background noise which is the acoustic sound waves from various sources that disrupt the original signal. Adaptive filters are found to be very reliable and effective in speech enhancement. However, one major problem is training. For training, we must have a priori knowledge of the noise-free speech. But practically, it is not possible to have a priori knowledge of the speech that has been recorded in a particular background. The work here proposes an adaptive filter-based algorithm for enhancement of the noisy speech. The only requirement is any clean short voice record of the same speaker. The work proposed here increases the output SNR by around 1.5 times the input SNR and also enhances the listening comfort of the speech.

Single‐microphone speech enhancement algorithms by using nonnegative matrix factorization can only utilize the temporal and spectral diversity of the received signal, making the performance of the noise suppression degrade rapidly in a complex environment. Microphone arrays have spatial selection and high signal gain, so it applies to the adverse noise conditions. In this paper, we present a new algorithm for speech enhancement based on two microphones with nonnegative matrix factorization. The interchannel characteristic of each nonnegative matrix factorization basis can be modeled by the adopted method, such as the amplitude ratios and the phase differences between channels. The results of the experiment confirm that the proposed algorithm is superior to other dual‐microphone speech enhancement algorithms.

A fast adaptive algorithm is developed for training optimal weighted order statistic (WOS) filters under the mean absolute error (MAE) criterion. This algorithm is implemented without use of threshold decomposition and can be applied to arbitrary real-valued signals directly. It can easily be extended to finite impulse response (FIR)-WOS hybrid filters. An interesting application of adaptive WOS and FIR WOS hybrid filters is presented. >

Speech enables easy human-to-human communication as well as human-to-machine interaction. However, the quality of speech degrades due to background noise in the environment, such as drone noise embedded in speech during search and rescue operations. Similarly, helicopter noise, airplane noise, and station noise reduce the quality of speech. Speech enhancement algorithms reduce background noise, resulting in a crystal clear and noise-free conversation. For many applications, it is also necessary to process these noisy speech signals at the edge node level. Thus, we propose implicit Wiener filter-based algorithm for speech enhancement using edge computing system. In the proposed algorithm, a first order recursive equation is used to estimate the noise. The performance of the proposed algorithm is evaluated for two speech utterances, one uttered by a male speaker and the other by a female speaker. Both utterances are degraded by different types of non-stationary noises such as exhibition, station, drone, helicopter, airplane, and white Gaussian stationary noise with different signal-to-noise ratios. Further, we compare the performance of the proposed speech enhancement algorithm with the conventional spectral subtraction algorithm. Performance evaluations using objective speech quality measures demonstrate that the proposed speech enhancement algorithm outperforms the spectral subtraction algorithm in estimating the clean speech from the noisy speech. Finally, we implement the proposed speech enhancement algorithm, in addition to the spectral subtraction algorithm, on the Raspberry Pi 4 Model B, which is a low power edge computing device.

This paper proposes novel heuristic algorithm called gravitational search algorithm (GSA) to speech enhancement. Stochastic and heuristic algorithms like particle swarm optimization (PSO) and some of its variants have been adapted to the field of speech enhancement in recent years. Although standard PSO (SPSO) finds good solutions, it suffers from the premature convergence by getting trapped into local optimum. In order to increase the diversity in search space and to improve the local searching capability, another recently developed heuristic algorithm GSA is proposed to speech enhancement in this paper. GSA is mainly constructed on the basis of law of gravity and the notion of mass interactions .The proposed algorithm is studied for real world noise condition called babble noise, at three different input SNR levels. To the best of our knowledge, there is no analysis about the intelligibility of enhanced speech using optimization techniques. In the present study, the proposed algorithm is compared with the standard PSO (SPSO) algorithm for dual channel speech enhancement, and the intelligibility analysis is also reported. Simulation results indicate that GSA-based algorithm outperforms the particle swarm optimization in adaptive noise cancellation with an improved speech quality and intelligibility.

Fast square-root RLS adaptive filtering algorithms