Reverberant Component Research Articles

An adaptive multichannel speech dereverberation approach based on robust estimation

Room reverberation usually produces adverse influence on the quality of speech in human-machine interaction scenarios. To improve the clarity and intelligibility of reverberant speech, an adaptive multichannel speech dereverberation approach is proposed in this work. Unlike the traditional least-squares cost function for speech dereverberation, a robust estimator with an adaptive parameter is adopted to define a flexible cost function of the adaptive speech dereverberation algorithm. As a consequence, the developed adaptive filtering algorithm can dynamically track the non-stationarity and non-Gaussian properties of speech so that the reverberation component of speech is effectively suppressed. Numerical simulation experiments validate the dereverberation performance of the proposed approach.

A tensor decomposition based multichannel linear prediction approach to speech dereverberation

Dereverberation technology is needed in a wide range of speech applications as reverberation often greatly degrades the quality and intelligibility of the speech signal of interest captured by microphones. The commonly used weighted-prediction-error method generally requires long prediction-error filters to remove the reverberation components, which makes it computationally expensive. To deal with this issue, this paper proposes a computationally efficient dereverberation algorithm based on tensor decomposition in which the long prediction-error filter is decomposed into a group of short sub-filters through multiple Kronecker products. Consequently, the high dimensional cross-correlation matrix that needs to be inverted in the dereverberation algorithm is then converted into a set of low dimensional matrices, which leads to significant reduction in the computational complexity. Simulation results demonstrate the properties of the proposed algorithm.

Crossband Filtering for Weighted Prediction Error-Based Speech Dereverberation

Weighted prediction error (WPE) is a linear prediction-based method extensively used to predict and attenuate the late reverberation component of an observed speech signal. This paper introduces an extended version of the WPE method to enhance the modeling accuracy in the time–frequency domain by incorporating crossband filters. Two approaches to extending the WPE while considering crossband filters are proposed and investigated. The first approach improves the model’s accuracy. However, it increases the computational complexity, while the second approach maintains the same computational complexity as the conventional WPE while still achieving improved accuracy and comparable performance to the first approach. To validate the effectiveness of the proposed methods, extensive simulations are conducted. The experimental results demonstrate that both methods outperform the conventional WPE regarding dereverberation performance. These findings highlight the potential of incorporating crossband filters in improving the accuracy and efficacy of the WPE method for dereverberation tasks.

Model based spectral feature analysis in reverberant acoustic fields

The spectral features of a reverberant acoustic field that can lead to improved source localization methods are investigated. Of particular interest are identification of deterministic and probabilistic features that characterize the environment and models of the acoustic transfer function that generate such features. For example, models that predict Schroeder’s invariant standard deviation in the pressure response at source-receiver distances where the direct sound approaches the reverberant field energy are proposed. Models examined include non-stationary Gaussian processes for the acoustic reverberant components and all-pass parametric models of the acoustic impulse response.

Accuracy of geometrical formula for predicting the insertion loss of balconies in two dimensions

A geometrical formula which includes the effect of first order reflections, diffraction, and a diffuse reverberant field is presented to predict the acoustic insertion loss of rectangular balconies on a building facade. This method is compared to two-dimensional boundary element predictions for numerous geometrical configurations. Results are presented as a function of the height of the shadow zone on the facade due to varying source position, with various balcony depths investigated. With a reflective ceiling or solid balustrade present, the effect of the modes is shown to make the diffuse assumption inaccurate at low frequencies. The frequency of each mode is compared to the analytical solution for the quasinormal modes of a baffled rectangular cavity, finding that this approximation is effective to predict the frequencies of the balcony modes. The understanding of the contribution of the direct and reverberant components inside the balcony cavity offers an explanation to the possibility of decreasing insertion loss with elevation angle which has been observed in three-dimensional scale model experiments in the literature and contradicts the insertion loss values recommended by ISO 12354-3:2017.

Weighted pressure matching with windowed targets for personal sound zones.

Personal sound zones (PSZ) systems use an array of loudspeakers to render independent audio signals to multiple listeners within a room. The performance of a PSZ system, designed using weighted pressure matching, depends on the selected target responses for the bright zone. In reverberant environments, the target responses are generally chosen to be the room impulse responses from one of the loudspeakers to the control points in the selected bright zone. This approach synthesizes the direct propagation component and all the reverberant components in the bright zone, while minimizing the energy in the dark zone. We present a theoretical analysis to show that high energy differences cannot be achieved for the diffuse reverberant components in the bright and dark zones, and so trying to synthesize these components in the bright zone does not lead to the best performance. It is then shown that the performance can be improved by using windowed versions of these measured impulse responses as target signals, in order to control which reverberant components are synthesized in the bright zone and which are not. This observation is supported by experimental measurements in two scenarios with different levels of reverberation.

VACE-WPE: Virtual Acoustic Channel Expansion Based on Neural Networks for Weighted Prediction Error-Based Speech Dereverberation

Speech dereverberation is an important issue for many real-world speech processing applications. Among the techniques developed, the weighted prediction error (WPE) algorithm has been widely adopted and advanced over the last decade, which blindly cancels out the late reverberation component from the reverberant mixture of microphone signals. In this study, we extend the neural-network-based virtual acoustic channel expansion (VACE) framework for the WPE-based speech dereverberation, a variant of the WPE that we recently proposed to enable the use of dual-channel WPE algorithm in a single-microphone speech dereverberation scenario. Based on the previous study, some ablation studies are conducted regarding the constituents of the VACE-WPE in an offline processing scenario. These studies reveal the characteristics of the system, thereby simplifying the architecture and leading to the introduction of new strategies for training the neural network for the VACE. Experimental results demonstrate that VACE-WPE (our PyTorch implementation and pre-trained models are available from <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/dreadbird06/vace_wpe</uri> ) considerably outperforms its single-channel counterpart in simulated noisy reverberant environments in terms of objective speech quality and is superior to the single-channel WPE as well as several fully neural speech dereverberation methods when employed as the front-end for the far-field automatic speech recognizer.

Improved spatialization performance for joint speech dereverberation and noise reduction in binaural hearing aids

ObjectiveIn this work we propose a unified solution for joint speech dereverberation, noise reduction and spatial preservation of either diffuse noise fields or single-point noise sources in binaural hearing aid applications, which is based on the multichannel Wiener filter (MWF) with preservation of the interaural coherence (MWF-IC) method. MethodsThe original interaural coherence (IC) auxiliary penalty cost function in the MWF-IC, which comprises information about both interfering noise and late reverberation component, is modified to consider the interference noise information only. It provides a new regularization form for preserving both interaural time difference (ITD) and IC binaural cues of the residual noise. This approach is theoretically supported by recent findings about the relationship between IC and ITD binaural cues for point acoustic sources. ResultsExtensive simulation results with objective measures, as well as psychoacoustic experiments with normal hearing volunteers indicate that the proposed method outperforms the preservation of spatial localization cues of its original version, keeping approximately the same performance in terms of noise-reduction and speech-dereverberation. ConclusionThe proposed scheme expands the MWF-IC original spatial cue preservation capability from only diffuse noise fields to either diffuse noise fields or a single-point noise source. SignificanceThis is the first MWF-based speech dereverberation and noise reduction method for binaural hearing aid that provides perceptually relevant preservation of acoustic scenarios comprised of one speech source and either a single-point noise source or a diffuse noise field.

Blind Signal Dereverberation Based on Mixture of Weighted Prediction Error Models

We extend the linear prediction-based dereverberation method called weighted prediction error (WPE). WPE optimizes a causal finite impulse response (FIR) filter that predicts the late reverberation components of an observed signal. However, by the multi-input/output inverse (MINT) theorem, in general, such FIR filters exist only when the number of sources is fewer than that of the microphones and no ambient noise exists. To mitigate the model error of WPE in adverse environments, we propose a mixture model of multiple WPEs in which the time frames are divided into clusters in each frequency bin and the WPE's causal FIR filter is optimized in each cluster. Experimental results show that our proposed method significantly improves the dereverberation performance of WPE when the noise level is high or the number of microphones is small.

Online Speech Dereverberation Using Mixture of Multichannel Linear Prediction Models

We extend the state-of-the-art online dereverberation method, online weighted prediction error (WPE), which predicts late reverberation components using a multichannel linear prediction (MCLP) filter. The multi-input/output inverse theorem states that in general such an MCLP filter for WPE exists only if the number of sources is less than that of microphones M and there is no additive noise; otherwise, the WPE model has error, degrading its performance especially when M is small. To mitigate this WPE drawback, we recently developed an offline dereverberation method called switching WPE (SwWPE), which has multiple MCLP filters and switches them for each time-frequency bin. We here propose a recursive least squares (RLS) algorithm for the online optimization of SwWPE. Our proposed RLS has forgetting weights for each MCLP filter, and their decay rates are controlled based on a switching mechanism of SwWPE to stably improve the dereverberation performance of the online WPE. Experimental results show that our online SwWPE clearly outperforms online WPE in speech dereverberation tasks.

Prefiltering Approach of Adaptive Eigenvalue Decomposition Method for an Improved Time Delay Estimation in Room Environment

Time delay estimation is an essential step in sound source localization and beamforming systems, and an extensive amount of research has been performed on this subject. This task entails the accurate estimation of the relative time delay between two microphone signals originating from the same source. To overcome the limitations of the existing methods, the adaptive eigenvalue decomposition (AED) algorithm was developed for time delay estimation in reverberant acoustic environments in the early 2000s. This paper attempts to improve delay estimation performance using autoregressive model-based prefiltering for the AED algorithm. The proposed method establishes an autoregressive model of the room impulse response beforehand. Then, the model is utilized as a linear prediction filter to remove the reverberation component from the microphone signals, which improves the estimation capability of the algorithm. Monte Carlo experiments are performed to demonstrate the improved performance for various reverberation levels and signal-to-noise ratios. Provided that the noise level is moderate, the proposed method is shown to greatly increase the accuracy of the conventional approach in a highly reverberant room.

Robust Dereverberation With Kronecker Product Based Multichannel Linear Prediction

Reverberation impairs not only the speech quality, but also intelligibility. The weighted-prediction-error (WPE) method, which estimates the late reverberation component based on a multichannel linear predictor, is by far one of the most effective algorithms for dereverberation. Generally, the WPE prediction filter in every short-time-Fourier-transform (STFT) subband has to be long enough to estimate accurately the late reverberation component. As a consequence, WPE is computationally expensive, which makes it difficult to implement into real-time embedded or edge computing devices. Moreover, WPE is sensitive to additive noise and its performance may suffer from dramatic degradation even in environments where the signal-to-noise ratio (SNR) is high. To address these drawbacks, this letter proposes to decompose the multichannel linear prediction filter as a Kronecker product of a temporal (interframe) prediction filter and a spatial filter. An iterative algorithm is then developed to optimize the two filters. In comparison with the original WPE algorithm, the presented method not only exhibits better performance in terms of dereverberation and robustness to additive noise, as there are fewer parameters to estimate for a given number of observation signal samples, but is also computationally more efficient, since the dimensions of the covariance matrices after Kronecker product decomposition are smaller.

Investigation of acoustic transmission beneath a railway vehicle by using statistical energy analysis and an equivalent source model

An approach is presented for modelling the noise propagation beneath the train floor and this is applied to rolling noise sources. It is assumed that the sound incident on the train floor is made up of a direct and a reverberant component. A combination of two numerical modelling approaches is considered to deal with these: an equivalent source model, to represent the direct component, and statistical energy analysis (SEA) for the reverberant part. In the equivalent source model, the wheel is replaced by monopole and dipole sources, which represent its radial and axial radiation. The rail vertical vibration and the sleepers are replaced by arrays of monopole sources while the rail lateral vibration is replaced by an array of lateral dipoles. The sound power of the rolling noise is obtained by using the TWINS model. In the SEA model, the region beneath the train floor is divided into several volumes and the power input to these subsystems is assumed to be due to the first reflections from the train floor and the ground. The reverberant and direct sound have very similar contributions to the total sound power incident on the train floor although this depends on how the equipment is arranged beneath the train. The modelling approach is verified by comparing the predicted sound pressure levels with laboratory measurements and with field tests.

Binaural Direct-to-Reverberant Energy Ratio and Speaker Distance Estimation

This article addresses the problem of distance estimation using binaural hearing aid microphones in reverberant rooms. Among several distance indicators, the direct-to-reverberant energy ratio (DRR) has been shown to be more effective than other features. Therefore, we present two novel approaches to estimate the DRR of binaural signals. The first method is based on the interaural magnitude-squared coherence whereas the second approach uses stochastic maximum likelihood beamforming to estimate the power of the direct and reverberant components. The proposed DRR estimation algorithms are integrated into a distance estimation technique. When based solely on DRR, the distance estimation algorithm requires calibration where naturally the critical distance is a good calibration point. We thus propose two approaches for the calibration of the distance estimation algorithm: Informed calibration using the critical distance of the reverberant room and blind calibration using the listener's own voice. Results across various acoustical environments show the benefit of the proposed algorithms for the estimation of sound source distances up to 3 m with an estimation error of about 35 cm using informed calibration and about 1 m using the fully blind calibration strategy.

Simulations and Subjective Rating of Acoustic Conditions in a Symphony Orchestra—A Case Study

Acoustic conditions in a symphony orchestra on a concert hall stage are very different from those on an empty stage. Since inter-orchestral sound transmission and other acoustic conditions with the orchestra present is easier to simulate than to measure, a method for simulations in Odeon models of orchestras in different rooms was developed by this author. This method was applied in the Grieghallen Renewal Project, which involved changes in concert hall, orchestra pit, and rehearsal hall. The resident orchestra members gave their overall rating of playing conditions in the home venues in addition to a number of international venues. Acoustical conditions in the rated venues were simulated and compared with ratings. Several metrics were investigated, and their correlation with subjective ratings varied between r2 = 0.09 and r2 = 0.85. It turned out the orchestra clearly preferred to play in conditions where the direct component and the reverberant component of the inter-orchestral sound-transmission on average were equally strong; |D-R| = 0. Any deviation from equality was associated with reduced preference, with correlation coefficient r = −0.92. Several interesting implications and interpretations of the result are discussed in the paper.

Parameterization of the binaural room transfer function using modal decomposition.

Binaural room responses are normally measured on a listening subject in a room. The measurements, however, rapidly change with the source and receiver position. In addition, the measurements taken in a room can only be used to simulate scenes of that environment. In this work, an efficient parameterization of the binaural room transfer function is proposed, which has separable representations for the direct-path component and the reverberation component, thus providing a flexible way to generate binaural room responses for different environments and listeners. In addition, this parameterization uses wave equation solutions as basis functions and is continuous in space.

A Bayesian approach to convolutive nonnegative matrix factorization for blind speech dereverberation

Abstract When a signal is recorded in an enclosed room, it typically gets affected by reverberation. This degradation represents a problem when dealing with audio signals, particularly in the field of speech signal processing, such as automatic speech recognition. Although there are some approaches to deal with this issue that are quite satisfactory under certain conditions, constructing a method that works well in a general context still poses a significant challenge. In this article, we propose a Bayesian approach based on convolutive nonnegative matrix factorization that uses prior distributions in order to impose certain characteristics over the time-frequency components of the restored signal and the reverberant components. An algorithm for implementing the method is described and tested. Comparisons of the results against those obtained with state-of-the-art methods are presented, showing significant improvement.

A perceptually motivated LP residual estimator in noisy and reverberant environments

Both reverberation and additive noise can degrade the quality of recorded speech and thus should be suppressed simultaneously. Previous studies have shown that the generalized singular value decomposition (GSVD) has the capability of suppressing the additive noise effectively, but it is not often applied for speech dereverberation since reverberation is considered to be convolutive as well as colored noise. Recently, we revealed that late reverberation is also additive and relatively white interference component in the linear prediction (LP) residual domain. To suppress both late reverberation and additive noise, we have proposed an optimal filter for LP residual estimator (LPRE) based on a constrained minimum mean square error (CMMSE) by using GSVD in single channel speech enhancement, where the algorithm is referred as CMMSE-GSVD-LPRE. Experimental results have shown a better performance of the CMMSE-GSVD-LPRE than spectral subtraction methods, but some residual noise and reverberation components are still audible and annoying. To solve this problem, this paper incorporates the masking properties of the human auditory system in the LP residual domain to further suppress these residual noise and reverberation components while reducing speech distortion at the same time. Various simulation experiments are conducted, and the results show an improved performance of the proposed algorithm. Experimental results with speech recorded in noisy and reverberant environments further confirm the effectiveness of the proposed algorithm in real-world environments.

Robust Speech Dereverberation With a Neural Network-Based Post-Filter That Exploits Multi-Conditional Training of Binaural Cues

This study presents an algorithm for binaural speech dereverberation based on the supervised learning of short-term binaural cues. The proposed system combined a delay-and-sum beamformer with a neural network-based post-filter that attenuated reverberant components in individual time-frequency units. A multi-conditional training procedure was used to simulate the uncertainties of short-term binaural cues in response to room reverberation by mixing the direct part of head related impulse responses (HRIRs) with diffuse noise. Despite being trained with only anechoic HRIRs, the proposed dereverberation algorithm was tested in a variety of reverberant environments and achieved considerable improvements relative to a coherence-based approach in terms of three objective metrics reflecting speech quality and speech intelligibility. Moreover, a systematic evaluation showed that the proposed system generalized very well to a wide range of acoustic conditions, including various measured binaural room impulse responses reflecting different reverberation times, azimuth positions spanning the entire frontal hemifield, various source-receiver distances as well as different artificial heads.

Effects of wide dynamic range compression on speech signals with respect to reverberation

The sound perception in enclosed spaces is dominated by the room acoustics properties of the enclosures. Today, it is known that reverberation generated by walls and obstacles challenges hearing-impaired people, even with hearing aids, and several studies have been conducted to address this problem. However, relatively little is known about how various signal processing blocks (i.e., beamforming, wide dynamic range compression) within hearing aids affect the reverberation content of the speech signals. Aim of this work was to investigate and quantify the effects of wide dynamic range compression on the reverberant component of speech signals employing both subjective and objective methods. Several objective metrics which correlate with reverberation were applied on the speech signals before and after the compression. Moreover, a listening test with 14 normal hearing participants was performed to assess whether the changes in the reverberation content of the compressed signals were perceivable. The perceptual results show that the gain model changed the perception of reverberation of the speech signals tested. Finally, the correlation between the objective metrics and the perceptual results of the listening test was then investigated, indicating increased reverberation content for the compressed signals.