Speech enhancement (SE) systems aim to improve the quality and intelligibility of degraded speech signals obtained from far-field microphones. Subjective evaluation of the intelligibility performance of these SE systems is uncommon. Instead, objective intelligibility measures (OIMs) are generally used to predict subjective performance increases. Many recent deep learning (DL) based SE systems, are expected to improve the intelligibility of degraded speech as measured by OIMs. However, validation of the ability of these OIMs to predict subjective intelligibility when enhancing a speech signal using DL-based systems is lacking. Therefore, in this study, we evaluate the predictive performance of five popular OIMs. We compare the metrics' predictions with subjective results. For this purpose, we recruited 50 human listeners, and subjectively tested both single channel and multi-channel Deep Complex Convolutional Recurrent Network (DCCRN) based speech enhancement systems. We found that none of the OIMs gave reliable predictions, and that all OIMs overestimated the intelligibility of ‘enhanced’ speech signals.

This paper investigates the use of different room impulse response (RIR) simulation methods for synthesizing training data for deep neural network-based direction of arrival (DOA) estimation of speech in reverberant rooms.Different sets of synthetic RIRs are obtained using the image source method (ISM) and more advanced methods including diffuse reflections and/or source directivity. Multi-layer perceptron (MLP) deep neural network (DNN) models are trained on generalized cross correlation (GCC) features extracted for each set. Finally, models are tested on features obtained from measured RIRs.This study shows the importance of training with RIRs from directive sources, as resultant DOA models achieved up to 51% error reduction compared to the steered response power with phase transform (SRP-PHAT) baseline (significant with p << .01), while models trained with RIRs from omnidirectional sources did worse than the baseline. The performance difference was specifically present when estimating the azimuth of speakers not facing the array directly.

A channel sounder operating in the 60 GHz band is described. It is designed using off-the-shelf components and programmed using model based design. Results from indoor measurements are provided and discussed in the context of published channel models for similar environments. The results include LOS measurements from a narrow office corridor and mixed LOS and NLOS measurements from an open-space canteen.

In this paper, an experimental performance evaluation is carried out within a communication scenario that features two of the key enablers of 5G: (i) the use of post-OFDM modulations and (ii) an efficient use of the spectrum via spectrum sharing. The experimental lab set-up has been assembled so as to operate in conditions as realistic as possible via the actual realtime implementation of the involved transceivers and also via the utilization of propagation channels which have been recorded in a field measurement campaign and which are loaded in a channel emulator that also operates in real-time. The experimental results show that co-existence in a shared spectrum scenario is possible and that the performance degradation is kept at a low level as long as one of the two users is making use of spectrally agile post-OFDM modulations such as filterbank multicarrier (FBMC).

Hidden Markov models (HMMs) are powerful generative models for sequential data that have been used in automatic speech recognition for more than two decades. Despite their popularity, HMMs make inaccurate assumptions about speech signals, thereby limiting the achievable performance of the conventional speech recognizer. Penalized logistic regression (PLR) is a well-founded discriminative classifier with long roots in the history of statistics. Its classification performance is often compared with that of the popular support vector machine (SVM). However, for speech classification, only limited success with PLR has been reported, partially due to the difficulty with sequential data. In this paper, we present an elegant way of incorporating HMMs in the PLR framework. This leads to a powerful discriminative classifier that naturally handles sequential data. In this approach, speech classification is done using affine combinations of HMM log-likelihoods. We believe that such combinations of HMMs lead to a more accurate classifier than the conventional HMM-based classifier. Unlike similar approaches, we jointly estimate the HMM parameters and the PLR parameters using a single training criterion. The extension to continuous speech recognition is done via rescoring of N-best lists or lattices.

State-of-the-art pattern recognition approaches like neural networks or kernel methods have only had limited success in speech recognition. The difficulties often encountered include the varying lengths of speech signals as well as how to deal with sequences of labels (e.g., digit strings) and unknown segmentation. In this paper we present a combined hidden Markov model (HMM) and penalized logistic regression machine (PLRM) approach to continuous speech recognition that can cope with both of these difficulties. The key ingredients of our approach are N-best rescoring and PLRM with garbage class. Experiments on the Aurora2 connected digits database show significant increase in recognition accuracy relative to a purely HMM-based system.

In this work we investigate the use of a greedy training algorithm for use with the dual Penalized Logistic Regression Machine (dPLRM), and our target application is detection of broad class phonetic features. The use of a greedy training algorithm is meant to alleviate the infeasible memory and computational demands that arises during the learning phase when the amount of training data increases. We show that using only a subset of the training data, chosen in a greedy manner, we can achieve as good as or better performance as when using the full training set. We can also train dPLRMs using data sets that are significantly larger than what our current computational resources can accommodate when using non-greedy approaches.

We propose an algorithm for optimal clustering and nonuniform allocation of Gaussian kernels in scalar (feature) dimension to compress complex, Gaussian mixture-based, continuous density HMMs into computationally efficient, small footprint models. The symmetric Kullback-Leibler divergence (KLD) is used as the universal distortion measure and it is minimized in both kernel clustering and allocation procedures. The algorithm was tested on the resource management (RM) database. The original context-dependent HMMs can be compressed to any resolution, measured by the total number of clustered scalar kernel components. Good trade-offs between the recognition performance and model complexities have been obtained; the HMM can be compressed to 15-20% of the original model size, which needs 1-5% of multiplication/division operations, and results in almost negligible recognition performance degradation.

We propose an online filtering algorithm that aims to alleviate the decrease we see in ASR performance when the speech is corrupted by additive noise. Using an initial estimate of the noise distribution, the algorithm updates the noise model on a frame synchronous basis. Using Polyak averaging we obtain a sequence of robust, frame-synchronous noise model estimates, and a minimum mean square error (MMSE) filter is used to denoise the cepstral coefficients. The algorithm is compared to a batch version which uses several iterations of the EM-algorithm over the complete utterance to estimate the noise model, and it is shown that the performance obtained using the averaging of the noise model is comparable to the batch performance.

In this paper we present an approach that makes use of both Bayesian predictive classification (BPC) and parallel model combination (PMC) to achieve increased robustness towards noise. PMC provides a method for finding parameter estimates for speech corrupted by noise, while BPC is a method that com-pensates for uncertainty of parameter estimates. Thus, these methods can be combined in order to obtain knowledge about the mismatch situation and simultaneously account for uncer-tainty in this knowledge. We apply this technique in an unsu-pervised approach on the Aurora2 database and show that good performance is obtained. 1.

In our previous work (2003), we investigated a new approach to robust speech recognition. An exact procedure was developed to filter noisy cepstral coefficients in the mean-square-error sense, and it was shown that this method outperformed the well known vector Taylor series (VTS) approach, which in turn is based on linear approximations to the non-linear filtering problem. Unfortunately. the procedure presented involved several integral equations with no known closed form solution. Numerical integration techniques were needed, which in turn led to slow performance, and in some cases, numerical problems. In this work we address this problem by using piecewise approximations to the integrands, which in turn yield closed form solutions. The revised procedure is tested on a subset of the Aurora 2 database, and the results are compared with the original numerical integration based approach, as well as VTS.

We present an optimal clustering algorithm for grouping multivariate normal distributions into clusters using the divergence, a symmetric, information-theoretic distortion measure based on the Kullback-Liebler distance. Optimal solutions for normal distributions are shown to be obtained by solving a set of Riccati matrix equations and the optimal centroids are found by alternating the mean and covariance matrix intermediate solutions. The clustering performance of the new algorithm compared favorably against the conventional, non-optimal clustering solutions of sample mean and sample covariance in its overall rate-distortion and even distributions of samples across clusters. The resultant clusters were further tested on unsupervised adaptation of HMM parameters in a framework of structured maximum a posterior linear regression (SMAPLR). The Wall Street Journal database was used for the adaptation experiment. The recognition performance with respect to the word error rate, was significantly improved from a nonoptimal centroid (sample mean and covariance) of 32.6% to 27.6% and 27.5% for the diagonal and full covariance matrix cases, respectively.

ABSTRACTIn this paper we study various front-endfeatures, mod-eling and adaptation algorithms on the Aurora 3 databases,including auditory, moment, and AM-FMmodulation fea-tures, context-dependentdigit models, segmental K-meanstraining, discriminative training, and model adaptations.The evaluation results on Aurora 3 are presented with abrief summary of our Aurora 2 results.1. INTRODUCTIONThe Aurora evaluation is for researchers to test their algo-rithms on noise robustness and compare results measuredon the same databases. So far, there are two tasks on theAurora evaluation, Aurora 2 and 3, both are for connecteddigit recognition. While the Aurora 2 databases use thecontrolled experiments by adding noise digitally to cleanEnglish digit strings [1], the Aurora 3 databases are col-lected in a real-worldcar environment in 4 languages. Inthis paper, we report our evaluation results on two of thelanguages, Spanish and German.2. BELL LABS APPROACHESIn this section, we present our baseline system then describethe different feature sets that have been used for this eval-uation. Alternative training strategies and acoustic modeladaptation techniques are also reviewed.A. Context-DependentModel: Similar to last year ap-proach [1], we have decided to use context-dependent(CD)digit models, together with Bell Labs recognition engine asbackend. This contrasts with the officialAurora backendthat is based on whole-worddigit models and the HTK en-gine. The officialbackend setup typically leads to poorerresults, especially in larger databases, and we believe that abetter baseline is beneficialto properly study the effect ofdifferent front-endson the finalrecognition performance.Last year, we investigated several approaches to buildCD digit models. Given the limited amount of trainingdata, especially in the Aurora3 databases, it is required torely on some tying techniques to build CD digit models.The Head-Body-Tail digit model structure (HBT) assumesthat CD digit models are built by concatenating a left-context-dependentunit (head) with a context-independentunit(body)followedbyaright-context-dependentunit(tail).For example, assuming that the lexicon contains 10 digitsplus a silence model, each digit model consists of a set of 1body, 11 heads and 11 tails (representing all left/right con-texts) [2]. We typically model each head and tail with a3-stateHMM, while a 4-stateHMM is used for each body.Most of the experiments done this year have been based onthe HBT structure. CD digit models can also be built as tri-phone modelsusing a decision tree. This is the approach weintroduced last year [1], and some of this year experimentshave been carried out using this model topology.B. Auditory Feature: The new auditory front-endinour recognition system was developed to mimic the robusthuman hearing in adverse acoustic environments [3, 4]. Inthe front-end,efficientsignal processing functions were im-plemented to satisfy both real-timeand computation costrequirements. Based on the analysis of the outer and mid-dle ear, a transfer function was constructed to replace thecommonly used preemphasis filter, and then a new set ofdigital auditory filters,which simulate auditory filteringinthe cochlea, replaces those used in the MFCC and PLP.The auditory feature extraction procedure consists of: anouter-middle-eartransfer function, FFT, frequency conver-sion from linear to the Bark scale, auditory filtering,non-linearity,and discrete cosine transform (DCT). In our previ-ous study[3], the feature has been evaluated in two tasks:connected-digit and large vocabulary, continuous speechrecognitionundervariousnoiseconditions,usingbothhand-setand hands-freedatainlandlineand wirelesstransmissionwith additive car and babble noise. Compared with theLPCC, MFCC, MEL-LPCC,and PLP features, the audi-tory feature achieved significantperformance improvement

In this paper we present an approach to transformation based model adaptation that combines a fast, closed form solution to the MAP estimation of our transforms with robust priors. The robust priors are found using the technique of hierarchical priors, and a closed form solution is achieved by choosing diagonally constrained affine transformations and a suitable family of prior distributions for these transformations. We show that the method gives results comparable to other algorithms, but with significantly reduced computational complexity and memory demands. Experiments are conducted on the SI Recognition Outlier task from the Wall Street Journal corpus, where speaker independent models have to be adapted to handle speech from non-native speakers.