Speaker Verification Experiments Research Articles

Optimization of the parameters characterizing sigmoidal rate-level functions based on acoustic features

This paper describes the development of an optimal sigmoidal rate-level function that is a component of many models of the peripheral auditory system. The optimization makes use of a set of criteria defined exclusively on the basis of physical attributes of the input sound that are inspired by physiological evidence. The criteria developed attempt to discriminate between a degraded speech signal and noise to preserve the maximum amount of information in the linear region of the sigmoidal curve, and to minimize the effects of distortion in the saturating regions. The performance of the proposed optimal sigmoidal function is validated by text-independent speaker-verification experiments with signals corrupted by additive noise at different SNRs. The experimental results suggest that the approach presented in combination with cepstral variance normalization can lead to relative reductions in equal error rate as great as 40% when compared with the use of baseline MFCC coefficients for some SNRs.

VoxGrid: a mobile voice verification system

PurposeThe purpose of this paper is to introduce VoxGrid, a mobile voice verification system intended for improving the security of the username‐password authentication scheme.Design/methodology/approachThe system incorporates text‐dependant speaker verification via mobile devices that provides for a three‐factor authentication scheme for granting authorised access to certain websites or applications. The same speech recognition engine used by Google Voice Search is utilised to provide voice‐to‐text feature. All verification tasks are performed on a centralised server to minimise computing requirements on mobile platforms where feature extractions is executed using Mel Frequency Cepstral Coefficients. The resulting features are transmitted to the server instead of raw voice data to reduce network load. Actual voice verification takes place in the central server using Vector Quantisation.FindingsThe initial results have indicated that VoxGrid is capable of providing an additional level of security on user authentications at a low cost and without using extra security tokens other than one's voice with a good enough performance given the limited resources available during testing.Originality/valuePast speaker verification experiments have been conducted but we see that this is the first time it is done on mobile devices with a client‐server architecture using K‐Means Clustering and Vector Quantisation. Future improvements on performance and testing could result in a more secure mobile computing environment.

Low-Variance Multitaper MFCC Features: A Case Study in Robust Speaker Verification

In speech and audio applications, short-term signal spectrum is often represented using mel-frequency cepstral coefficients (MFCCs) computed from a windowed discrete Fourier transform (DFT). Windowing reduces spectral leakage but variance of the spectrum estimate remains high. An elegant extension to windowed DFT is the so-called multitaper method which uses multiple time-domain windows (tapers) with frequency-domain averaging. Multitapers have received little attention in speech processing even though they produce low-variance features. In this paper, we propose the multitaper method for MFCC extraction with a practical focus. We provide, first, detailed statistical analysis of MFCC bias and variance using autoregressive process simulations on the TIMIT corpus. For speaker verification experiments on the NIST 2002 and 2008 SRE corpora, we consider three Gaussian mixture model based classifiers with universal background model (GMM-UBM), support vector machine (GMM-SVM) and joint factor analysis (GMM-JFA). Multitapers improve MinDCF over the baseline windowed DFT by relative 20.4% (GMM-SVM) and 13.7% (GMM-JFA) on the interview-interview condition in NIST 2008. The GMM-JFA system further reduces MinDCF by 18.7% on the telephone data. With these improvements and generally noncritical parameter selection, multitaper MFCCs are a viable candidate for replacing the conventional MFCCs.

New Missing Features Mask Estimation Method for Speaker Recognition in Noisy Environments

RESUMEN Currently, many speaker recognition applications must handle speech corrupted by environmental additiv e noise without having a priori knowledge about the charact eristics of noise. Some previous works in speaker r ecognition have used Missing Feature (MF) approach to compensate for noise. In most of those applications the spe ctral reliability decision step is done using the Signal to Noise Ratio (SNR) criterion. This has the goal o f enhancing signal power rather than noise power, which could b e dangerous in speaker recognition tasks, because u seful speaker information could be removed. This work proposes a new mask estimation method based on Speaker Discriminative Information (SDI) for determining spectral reliabil ity in speaker recognition applications based on th e MF approach. The proposal was evaluated through speaker verification experiments in speech corrupted by additive no ise. Experiments demonstrated that this new criterion ha s a promising performance in speaker verification t asks.

Speaker Identification and Verification by Combining MFCC and Phase Information

In conventional speaker recognition methods based on Mel-frequency cepstral coefficients (MFCCs), phase information has hitherto been ignored. In this paper, we propose a phase information extraction method that normalizes the change variation in the phase according to the frame position of the input speech and combines the phase information with MFCCs in text-independent speaker identification and verification methods. There is a problem with the original phase information extraction method when comparing two phase values. For example, the difference in the two values of π-\mathtildeθ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> and \mathtildeθ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> =-π+\mathtildeθ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> is 2π-2\mathtildeθ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> . If \mathtildeθ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> ≈ 0, then the difference ≈ 2π, despite the two phases being very similar to one another. To address this problem, we map the phase into coordinates on a unit circle. Speaker identification and verification experiments are performed using the NTT database which consists of sentences uttered by 35 (22 male and 13 female) Japanese speakers with normal, fast and slow speaking modes during five sessions. Although the phase information-based method performs worse than the MFCC-based method, it augments the MFCC and the combination is useful for speaker recognition. The proposed modified phase information is more robust than the original phase information for all speaking modes. By integrating the modified phase information with the MFCCs, the speaker identification rate was improved to 98.8% from 97.4% (MFCC), and equal error rate for speaker verification was reduced to 0.45% from 0.72% (MFCC), respectively. We also conducted the speaker identification and verification experiments on a large-scale Japanese Newspaper Article Sentences (JNAS) database, a similar trend as NTT database was obtained.

Feature classification criterion for missing features mask estimation in robust speaker recognition

Currently, many speaker recognition applications must handle speech corrupted by environmental additive noise without having a priori knowledge about the characteristics of noise. Some previous works in speaker recognition have used the missing feature (MF) approach to compensate for noise. In most of those applications, the spectral reliability decision step is performed using the signal to noise ratio (SNR) criterion, which attempts to directly measure the relative signal to noise energy at each frequency. An alternative approach to spectral data reliability has been used with some success in the MF approach to speech recognition. Here, we compare the use of this new criterion with the SNR criterion for MF mask estimation in speaker recognition. The new reliability decision is based on the extraction and analysis of several spectro-temporal features from across the entire speech frame, but not across the time, which highlight the differences between spectral regions dominated by speech and by noise. We call it the feature classification (FC) criterion. It uses several spectral features to establish spectrogram reliability unlike SNR criterion that relies only in one feature: SNR. We evaluated our proposal through speaker verification experiments, in Ahumada speech database corrupted by different types of noise at various SNR levels. Experiments demonstrated that the FC criterion achieves considerably better recognition accuracy than the SNR criterion in the speaker verification tasks tested.

Speaker verification using excitation source information

In this work we develop a speaker recognition system based on the excitation source information and demonstrate its significance by comparing with the vocal tract information based system. The speaker-specific excitation information is extracted by the subsegmental, segmental and suprasegmental processing of the LP residual. The speaker-specific information from each level is modeled independently using Gaussian mixture modeling--universal background model (GMM-UBM) modeling and then combined at the score level. The significance of the proposed speaker recognition system is demonstrated by conducting speaker verification experiments on the NIST-03 database. Two different tests, namely, Clean test and Noisy test are conducted. In case of Clean test, the test speech signal is used as it is for verification. In case of Noisy test, the test speech is corrupted by factory noise (9 dB) and then used for verification. Even though for Clean test case, the proposed source based speaker recognition system still provides relatively poor performance than the vocal tract information, its performance is better for Noisy test case. Finally, for both clean and noisy cases, by providing different and robust speaker-specific evidences, the proposed system helps the vocal tract system to further improve the overall performance.

Using Jitter and Shimmer in speaker verification

Jitter and shimmer are measures of the fundamental frequency and amplitude cycle-to-cycle variations, respectively. Both features have been largely used for the description of pathological voices, and since they characterise some aspects concerning particular voices, they are expected to have a certain degree of speaker specificity. In the current work, jitter and shimmer are successfully used in a speaker verification experiment. Moreover, both measures are combined with spectral and prosodic features using several types of normalisation and fusion techniques in order to obtain better verification results. The overall speaker verification system is also improved by using histogram equalisation as a normalisation technique previous to fusing the features by support vector machines.

Using Kernel Discriminant Analysis to Improve the Characterization of the Alternative Hypothesis for Speaker Verification

Speaker verification can be viewed as a task of modeling and testing two hypotheses: the null hypothesis and the alternative hypothesis. Since the alternative hypothesis involves unknown impostors, it is usually hard to characterize a priori. In this paper, we propose improving the characterization of the alternative hypothesis by designing two decision functions based, respectively, on a weighted arithmetic combination and a weighted geometric combination of discriminative information derived from a set of pretrained background models. The parameters associated with the combinations are then optimized using two kernel discriminant analysis techniques, namely, the kernel Fisher discriminant (KFD) and support vector machine (SVM). The proposed approaches have two advantages over existing methods. The first is that they embed a trainable mechanism in the decision functions. The second is that they convert variable-length utterances into fixed-dimension characteristic vectors, which are easily processed by kernel discriminant analysis. The results of speaker-verification experiments conducted on two speech corpora show that the proposed methods outperform conventional likelihood ratio-based approaches.

Integration of Complementary Acoustic Features for Speaker Recognition

This letter describes a speaker verification system that uses complementary acoustic features derived from the vocal source excitation and the vocal tract system. A new feature set, named the wavelet octave coefficients of residues (WOCOR), is proposed to capture the spectro-temporal source excitation characteristics embedded in the linear predictive residual signal. WOCOR is used to supplement the conventional vocal tract-related features, in this case, the Mel-frequency cepstral coefficients (MFCC), for speaker verification. A novel confidence measure-based score fusion technique is applied to integrate WOCOR and MFCC. Speaker verification experiments are carried out on the NIST 2001 database. The equal error rate (EER) attained with the proposed method is 7.67%, in comparison to 9.30% of the conventional MFCC-based system

Improved Jacobian Adaptation for Robust Speaker Verification

Jacobian Adaptation (JA) has been successfully used in Automatic Speech Recognition (ASR) systems to adapt the acoustic models from the training to the testing noise conditions. In this work we present an improvement of JA for speaker verification, where a specific training noise reference is estimated for each speaker model. The new proposal, which will be referred to as Model-dependent Noise Reference Jacobian Adaptation (MNRJA), has consistently outperformed JA in our speaker verification experiments.

Speaker classification using composite hypothesis testing and list decoding

Speaker classification is seen as a hypothesis testing problem of J simple hypotheses and a composite hypothesis. The simple hypotheses represent target speakers while the composite hypothesis represents nontarget speakers. The simple hypotheses have well-defined distributions that are estimated from training signals. The distribution of the signal under the composite hypothesis is assumed to belong to a given family. The parameter of that distribution is assumed random with a prior distribution that is estimated from a large set of speakers. This formulation converts the problem to that of testing J+1 simple hypotheses. Signals corresponding to target and nontarget speakers are assumed Gaussian mixtures processes. Once the system has been trained, list decoding is applied in which a test signal is associated with a list of possible speakers. The probability that the correct speaker is on the list is maximized for a given average number of incorrect speakers on the list. Results from speaker identification and speaker verification experiments are reported. In speaker identification using a National Institute of Standards and Technology (NIST) database with 174 target speakers, over 77% correct identification was achieved for an average of less than two erroneous speakers on the list. Speaker verification experiments on a similar database yielded results, expressed in terms of the equal-error-rate, of 6.7% and 10.1% using two decision rules.

Speaker verification using sequence discriminant support vector machines

This paper presents a text-independent speaker verification system using support vector machines (SVMs) with score-space kernels. Score-space kernels generalize Fisher kernels and are based on underlying generative models such as Gaussian mixture models (GMMs). This approach provides direct discrimination between whole sequences, in contrast with the frame-level approaches at the heart of most current systems. The resultant SVMs have a very high dimensionality since it is related to the number of parameters in the underlying generative model. To address problems that arise in the resultant optimization we introduce a technique called spherical normalization that preconditions the Hessian matrix. We have performed speaker verification experiments using the PolyVar database. The SVM system presented here reduces the relative error rates by 34% compared to a GMM likelihood ratio system.

TEXT-INDEPENDENT SPEAKER VERIFICATION USING MINIMAL RESOURCE ALLOCATION NETWORKS

This paper presents a text-independent speaker verification system based on an online Radial Basis Function (RBF) network referred to as Minimal Resource Allocation Network (MRAN). MRAN is a sequential learning RBF, in which hidden neurons are added or removed as training progresses. LP-derived cepstral coefficients are used as feature vectors during training and verification phases. The performance of MRAN is compared with other well-known RBF and Elliptical Basis Function (EBF) based speaker verification methods in terms of error rates and computational complexity on a series of speaker verification experiments. The experiments use data from 258 speakers from the phonetically balancedcontinuous speech corpus TIMIT. The results show that MRAN produces comparable error rates to other methods with much less computational complexity.

A Fused Hidden Markov Model With Application to Bimodal Speech Processing

This paper presents a novel fused hidden Markov model (fused HMM) for integrating tightly coupled time series, such as audio and visual features of speech. In this model, the time series are first modeled by two conventional HMMs separately. The resulting HMMs are then fused together using a probabilistic fusion model, which is optimal according to the maximum entropy principle and a maximum mutual information criterion. Simulations and bimodal speaker verification experiments show that the proposed model can significantly reduce the recognition errors in noiseless or noisy environments.

Estimation of elliptical basis function parameters by the EM algorithm with application to speaker verification

This paper proposes to incorporate full covariance matrices into the radial basis function (RBF) networks and to use the expectation-maximization (EM) algorithm to estimate the basis function parameters. The resulting networks, referred to as elliptical basis function (EBF) networks, are evaluated through a series of text-independent speaker verification experiments involving 258 speakers from a phonetically balanced, continuous speech corpus (TIMIT).We propose a verification procedure using RBF and EBF networks as speaker models and show that the networks are readily applicable to verifying speakers using LP-derived cepstral coefficients as features. Experimental results show that small EBF networks with basis function parameters estimated by the EM algorithm outperform the large RBF networks trained in the conventional approach. The results also show that the equal error rate achieved by the EBF networks is about two-third of that achieved by the vetor quantization (VQ)-based speaker models.

A study of models anda priori threshold updating in speaker verification

This article explores a method for speaker verification to appropriately set the threshold value used to judge the identities of individual speakers and also considers methods to update speaker models and the robustness of the speaker models so as to make the models more robust with respect to utterance variations, using a small amount of data for updating that was recently uttered. The speaker model is represented by the hidden Markov model (HMM). In the model updating, the parameters of the speaker HMM are estimated from the data for updating and the current parameter values. For setting the threshold, the new threshold value for each speaker is set with the initial value as a value that is passed to an FA rate higher than the equal error rate (a value for which the false rejection rate and the false acceptance rate are equal), which is calculated from the data for updating and which steadily approaches the value that was passed to the equal error rate in concert with updating of the speaker HMM. The results of evaluating this method using text-dependent and text-prompted speaker verification experiments with twenty speakers shows that the average error rate fell by roughly 40% for the text-independent type and by roughly 80% for the text-prompted type as compared to when the model and threshold value are not updated. © 1999 Scripta Technica, Syst Comp Jpn, 30(13): 96–105, 1999

A study of models and a priori threshold updating in speaker verification

This article explores a method for speaker verification to appropriately set the threshold value used to judge the identities of individual speakers and also considers methods to update speaker models and the robustness of the speaker models so as to make the models more robust with respect to utterance variations, using a small amount of data for updating that was recently uttered. The speaker model is represented by the hidden Markov model (HMM). In the model updating, the parameters of the speaker HMM are estimated from the data for updating and the current parameter values. For setting the threshold, the new threshold value for each speaker is set with the initial value as a value that is passed to an FA rate higher than the equal error rate (a value for which the false rejection rate and the false acceptance rate are equal), which is calculated from the data for updating and which steadily approaches the value that was passed to the equal error rate in concert with updating of the speaker HMM. The results of evaluating this method using text-dependent and text-prompted speaker verification experiments with twenty speakers shows that the average error rate fell by roughly 40% for the text-independent type and by roughly 80% for the text-prompted type as compared to when the model and threshold value are not updated. © 1999 Scripta Technica, Syst Comp Jpn, 30(13): 96–105, 1999

Speaker recognition using neural networks and conventional classifiers

An evaluation of various classifiers for text-independent speaker recognition is presented. In addition, a new classifier is examined for this application. The new classifier is called the modified neural tree network (MNTN). The MNTN is a hierarchical classifier that combines the properties of decision trees and feedforward neural networks. The MNTN differs from the standard NTN in both the new learning rule used and the pruning criteria. The MNTN is evaluated for several speaker recognition experiments. These include closed- and open-set speaker identification and speaker verification. The database used is a subset of the TIMIT database consisting of 38 speakers from the same dialect region. The MNTN is compared with nearest neighbor classifiers, full-search, and tree-structured vector quantization (VQ) classifiers, multilayer perceptrons (MLPs), and decision trees. For closed-set speaker identification experiments, the full-search VQ classifier and MNTN demonstrate comparable performance. Both methods perform significantly better than the other classifiers for this task. The MNTN and full-search VQ classifiers are also compared for several speaker verification and open-set speaker-identification experiments. The MNTN is found to perform better than full-search VQ classifiers for both of these applications. In addition to matching or exceeding the performance of the VQ classifier for these applications, the MNTN also provides a logarithmic saving for retrieval.

HMM speaker verification with sparse training data on telephone quality speech

Speaker verification experiments using discrete and semi-continuous HMMs with telephone quality isolated digits are reported. The models were trained with varying numbers of tokens, giving equal error rates of 14% and 12%, respectively, on single isolated digits, and 4% and 2% on a sequence of 12 isolated digits.