Gaussian Mixture Model Supervector Research Articles

Speaker Identification Analysis Based on Long-Term Acoustic Characteristics with Minimal Performance

The identity of the speakers depends on the phonological properties acquired from the speech. The Mel-Frequency Cepstral Coefficients (MFCC) are better researched for derived the acoustic characteristic. This speaker model is based on a small representation and the characteristics of the acoustic features. These are derived from the speaker model and the cartographic representation by the MFCCs. The MFCC is used for independent monitoring of speaker text. There is a problem with the recognition of speakers by small representation, so proposed the Gaussian Mixture Model (GMM), mean super vector core for training. Unknown vector modules are cleared using rarity and experiments based on the TMIT database. The I-vector algorithm is proposed for the effective improvement of ASR (Automatic Speaker Recognition). The Atom Aligned Sparse Representation (AASR) is used to describe the speaker-based model. The Short Representation Classification (SRC) is used to describe the speaker recognition report. A robust short coding is based on the Maximum Likelihood Estimation (MIE) to clarify the problem in small representation. Strong speaker verification based on a small representation of GMM super vectors. Strong speaker verification based on a small representation of GMM super vectors.

Model Identifikasi Pemalsuan Ijazah menggunakan Gabor Wavelet dan Gaussian Mixture Models Super Vektor (GMM-SV)

Various cases occur related to certificate falsification and some people and educational institutions have to deal with the law, this problem is not impossible to abuse along with advances and technological innovation with various tools that can be used by anyone. Identifying the diploma document must be of particular concern to tertiary institutions to minimize the associated fake diplomas and the diploma legalization process. In legalizing the diploma for STMIK Indonesia Banda Aceh students, checking the authenticity of the certificate is only by bringing the original certificate and photocopy of the certificate or by contacting the academic party who issued the certificate, this process is sometimes missed by officers when the queue is crowded. The specific objectives of the research include implementing a model and feature method of Gabor Wavelet and Gaussian Mixture Models Super Vector (GMM-SV) for document identification to speed up diploma identification. The flow of this research starts from the input in the form of a basic image as an image that a reference for the authenticity of the diploma. Then the test image input is an image that will be tested for authenticity. The results showed that using the Gabor Wavelet feature and the Gaussian Mixture Models Super Vector (GMM-SV) could identify fake diplomas with an accuracy rate of 92.8%.Keywords:Model, Identification, Certificate Falsification, Gabor Wavelet, Gaussian Mixture Models Super Vector (GMM-SV).

Speaker age and gender classification using GMM supervector and NAP channel compensation method

One of the most important factors affecting the performance of speech-based recognition systems is the differences between training and test conditions. The Nuisance attribute projection (NAP) is an effective method for eliminating these differences, called channel effects. In this study, the effects of the NAP approach in determining age and gender groups are investigated. Mel-frequency cepstral coefficients and delta coefficients are used as a feature and Gaussian mixture models (GMM) adapted from the universal background model by maximum-a-posteriori method are used for the modeling of age and gender classes. After the GMMs corresponding to each speech are converted into mean supervectors, they are applied to a Support Vector Machine (SVM), and speeches are classified according to the age and gender group of the speakers. While linear GMM kernel based on Kullback–Leibler divergence is used instead of standard SVM kernels, the NAP channel subspace size is changed between 20 and 200 and the number of GMM components is changed between 32 and 512 to determine the optimum values for these parameters. In the tests on the aGender database, the optimum number of components is determined as 128, and the optimum NAP channel subspace size is determined as 45. The age and gender classification accuracy of the system, which is developed using these optimum parameters, is increased from 60.52 to 62.03% with the use of NAP. In addition, age classification accuracy is increased from 60.23 to 61.82% and gender classification accuracy is increased from 91.71 to 92.30%.

Adversarially Learned Total Variability Embedding for Speaker Recognition with Random Digit Strings.

Over the recent years, various research has been conducted to investigate methods for verifying users with a short randomized pass-phrase due to the increasing demand for voice-based authentication systems. In this paper, we propose a novel technique for extracting an i-vector-like feature based on an adversarially learned inference (ALI) model which summarizes the variability within the Gaussian mixture model (GMM) distribution through a nonlinear process. Analogous to the previously proposed variational autoencoder (VAE)-based feature extractor, the proposed ALI-based model is trained to generate the GMM supervector according to the maximum likelihood criterion given the Baum–Welch statistics of the input utterance. However, to prevent the potential loss of information caused by the Kullback–Leibler divergence (KL divergence) regularization adopted in the VAE-based model training, the newly proposed ALI-based feature extractor exploits a joint discriminator to ensure that the generated latent variable and the GMM supervector are more realistic. The proposed framework is compared with the conventional i-vector and VAE-based methods using the TIDIGITS dataset. Experimental results show that the proposed method can represent the uncertainty caused by the short duration better than the VAE-based method. Furthermore, the proposed approach has shown great performance when applied in association with the standard i-vector framework.

Combination of Long-Term and Short-Term Features for Age Identification from Voice

In this paper, we propose to use Gaussian mixture model (GMM) supervectors in a feed-forward deep neural network (DNN) for age identification from voice. The GMM is trained with short- ...

Improving the Accuracy and Training Speed of Motor Imagery Brain-Computer Interfaces Using Wavelet-Based Combined Feature Vectors and Gaussian Mixture Model-Supervectors.

In this paper, we propose a set of wavelet-based combined feature vectors and a Gaussian mixture model (GMM)-supervector to enhance training speed and classification accuracy in motor imagery brain–computer interfaces. The proposed method is configured as follows: first, wavelet transforms are applied to extract the feature vectors for identification of motor imagery electroencephalography (EEG) and principal component analyses are used to reduce the dimensionality of the feature vectors and linearly combine them. Subsequently, the GMM universal background model is trained by the expectation–maximization (EM) algorithm to purify the training data and reduce its size. Finally, a purified and reduced GMM-supervector is used to train the support vector machine classifier. The performance of the proposed method was evaluated for three different motor imagery datasets in terms of accuracy, kappa, mutual information, and computation time, and compared with the state-of-the-art algorithms. The results from the study indicate that the proposed method achieves high accuracy with a small amount of training data compared with the state-of-the-art algorithms in motor imagery EEG classification.

Restricted Boltzmann machines for vector representation of speech in speaker recognition

Over the last few years, i-vectors have been the state-of-the-art technique in speaker recognition. Recent advances in Deep Learning (DL) technology have improved the quality of i-vectors but the DL techniques in use are computationally expensive and need phonetically labeled background data. The aim of this work is to develop an efficient alternative vector representation of speech by keeping the computational cost as low as possible and avoiding phonetic labels, which are not always accessible. The proposed vectors will be based on both Gaussian Mixture Models (GMM) and Restricted Boltzmann Machines (RBM) and will be referred to as GMM–RBM vectors. The role of RBM is to learn the total speaker and session variability among background GMM supervectors. This RBM, which will be referred to as Universal RBM (URBM), will then be used to transform unseen supervectors to the proposed low dimensional vectors. The use of different activation functions for training the URBM and different transformation functions for extracting the proposed vectors are investigated. At the end, a variant of Rectified Linear Units (ReLU) which is referred to as variable ReLU (VReLU) is proposed. Experiments on the core test condition 5 of NIST SRE 2010 show that comparable results with conventional i-vectors are achieved with a clearly lower computational load in the vector extraction process.

Identification of Reconstructed Speech

Both voice conversion and hidden Markov model-- (HMM) based speech synthesis can be used to produce artificial voices of a target speaker. They have shown great negative impacts on speaker verification (SV) systems. In order to enhance the security of SV systems, the techniques to detect converted/synthesized speech should be taken into consideration. During voice conversion and HMM-based synthesis, speech reconstruction is applied to transform a set of acoustic parameters to reconstructed speech. Hence, the identification of reconstructed speech can be used to distinguish converted/synthesized speech from human speech. Several related works on such identification have been reported. The equal error rates (EERs) lower than 5% of detecting reconstructed speech have been achieved. However, through the cross-database evaluations on different speech databases, we find that the EERs of several testing cases are higher than 10%. The robustness of detection algorithms to different speech databases needs to be improved. In this article, we propose an algorithm to identify the reconstructed speech. Three different speech databases and two different reconstruction methods are considered in our work, which has not been addressed in the reported works. The high-dimensional data visualization approach is used to analyze the effect of speech reconstruction on Mel-frequency cepstral coefficients (MFCC) of speech signals. The Gaussian mixture model supervectors of MFCC are used as acoustic features. Furthermore, a set of commonly used classification algorithms are applied to identify reconstructed speech. According to the comparison among different classification methods, linear discriminant analysis-ensemble classifiers are chosen in our algorithm. Extensive experimental results show that the EERs lower than 1% can be achieved by the proposed algorithm in most cases, outperforming the reported state-of-the-art identification techniques.

A study on the roles of total variability space and session variability modeling in speaker recognition

Speaker verification (SV) using i-vector concept becomes state-of-the-art. In this technique, speakers are projected onto the total variability space and represented by vectors called i-vectors. During testing, the i-vectors of the test speech segment and claimant are conditioned to compensate for the session variability before scoring. So, i-vector system can be viewed as two processing blocks: one is total variability space and the other is post-processing module. Several questions arise, such as, (i) which part of the i-vector system plays a major role in speaker verification: total variability space or post-processing task; (ii) is the post-processing module intrinsic to the total variability space? The motivation of this paper is to partially answer these questions by proposing several simpler speaker characterization systems for speaker verification, where speakers are represented by their speaker characterization vectors (SCVs). The SCVs are obtained by uniform segmentation of the speakers gaussian mixture models (GMMs)- and maximum likelihood linear regression (MLLR) super-vectors. We consider two adaptation approaches for GMM super-vector: one is maximum a posteriori and other is MLLR. Similarly to the i-vector, SCVs are post-processed for session variability compensation during testing. The proposed system shows promising performance when compared to the classical i-vector system which indicates that the post-processing task plays an major role in i-vector based SV system and is not intrinsic to the total variability space. All experimental results are shown on NIST 2008 SRE core condition.

Fast Scoring of Full Posterior PLDA Models

A low-dimensional representation of a speech segment, the so-called i-vector, in combination with probabilistic linear discriminant analysis (PLDA) models, is the current state-of-the-art in speaker recognition. An i-vector is a compact representation of a Gaussian Mixture Model (GMM) supervector, which captures most of the GMM supervectors variability. It is usually obtained by a MAP estimate of the mean of a posterior distribution. A new PLDA model has been recently presented that, unlike the standard one, exploits the intrinsic i-vector uncertainty. This approach, referred to in this paper as Full Posterior Distribution PLDA (FP-PLDA), is particularly effective for speaker detection of short and variable duration speech segments. It is, however, computationally far more expensive than standard PLDA, making it unattractive for real applications. This paper presents three simplifications of FP-PLDA based on approximate diagonalizations of matrices involved in FP-PLDA scoring. Using in sequence these approximations allows obtaining computational costs comparable to PLDA models, with only a small performance degradation with respect to the more accurate, but less efficient, FP-PLDA models. In particular, up to 10% better performance than PLDA is obtained, with similar computational complexity, on short speech segments of variable duration, randomly extracted from the interviews and telephone conversations included in the NIST SRE 2010 extended dataset. The benefits of the proposed diagonalization approaches have also been confirmed on a short utterance text-independent verification task, where approximately 43% and 34% improvement of the EER and minimum DCF08, respectively, has been obtained with respect to PLDA.

A Modified Adaptive GMM Approach Based GMM Supervector and I-vector Using NMF Decomposition for Robust Speaker Verification

We propose a new method to enhance performance of speaker verification by investigating a novel modification of adaptive Gaussian Mixture Model (GMM) training. This model is trained using a modified Expectation Maximization (EM) algorithm, combined with a modified Maximum A Posteriori (MAP) estimation based weight factor of observation probabilities, called the observation confidence. The observation confidence is calculated based on the SNR estimation. Based on this modified adaptive GMM training algorithm, we propose to construct GMM supervectors and i-vectors, which are considered as input feature vectors for SVM. Besides, the discriminant features for speaker verification are also exploited by using non-negative matrix factorization (NMF) in the GMM-supervector and i-vector space. Experiment results on utterances from Korean drama (“You came from the stars”) show that our proposed methods significantly outperform the baseline GMM-UBM, GMM-supervector and i-vector based SVM under various noisy conditions.

On robustness of speech based biometric systems against voice conversion attack

Voice conversion (VC) approach, which morphs the voice of a source speaker to be perceived as spoken by a specified target speaker, can be intentionally used to deceive the speaker identification (SID) and speaker verification (SV) systems that use speech biometric. Voice conversion spoofing attacks to imitate a particular speaker pose potential threat to these kinds of systems. In this paper, we first present an experimental study to evaluate the robustness of such systems against voice conversion disguise. We use Gaussian mixture model (GMM) based SID systems, GMM with universal background model (GMM-UBM) based SV systems and GMM supervector with support vector machine (GMM-SVM) based SV systems for this. Voice conversion is conducted by using three different techniques: GMM based VC technique, weighted frequency warping (WFW) based conversion method and its variation, where energy correction is disabled (WFW−). Evaluation is done by using intra-gender and cross-gender voice conversions between fifty male and fifty female speakers taken from TIMIT database. The result is indicated by degradation in the percentage of correct identification (POC) score in SID systems and degradation in equal error rate (EER) in all SV systems. Experimental results show that the GMM-SVM SV systems are more resilient against voice conversion spoofing attacks than GMM-UBM SV systems and all SID and SV systems are most vulnerable towards GMM based conversion than WFW and WFW− based conversion. From the results, it can also be said that, in general terms, all SID and SV systems are slightly more robust to voices converted through cross-gender conversion than intra-gender conversion. This work extended the study to find out the relationship between VC objective score and SV system performance in CMU ARCTIC database, which is a parallel corpus. The results of this experiment show an approach on quantifying objective score of voice conversion that can be related to the ability to spoof an SV system.

Relevance factor of maximum a posteriori adaptation for GMM–NAP–SVM in speaker and language recognition

This paper studies the relevance factor in maximum a posteriori (MAP) adaptation of Gaussian mixture model (GMM) for speaker and language recognition. Knowing that relevance factor determines how much the observed training data influence the model adaptation, thus the resulting GMM model, it is believed that more effective modeling can be achieved if the relevance factor is adaptive to the corresponding data. We therefore provide a mathematic derivation for the estimation of relevance factor. GMM supervector support vector machine (SVM) with nuisance attribute projection (NAP) (GMM–NAP–SVM) has been reported to be effective and reliable for speaker and language recognition. Being a discriminative classifier in nature, a GMM–NAP–SVM system is sensitive to the magnitude and direction of a supervector in the high dimensional space. However, when characterizing a speech utterance with GMM supervector estimated through MAP, we observe that the resulting supervector is undesirably affected by the varying duration of the utterance. We propose an adaptive relevance factor that adapts to the duration to mitigate the variability effect due to the length of utterance. We give a systematic investigation on different types of relevance factor of MAP in different applicatively platforms. We show the efficacy of the data-dependent as well as adaptive relevance factors on the National Institute of Standards and Technology (NIST) speaker recognition evaluation (SRE) 2008 and language recognition evaluation (LRE) 2009 and 2011 tasks respectively.

Reducing computation in an i-vector speaker recognition system using a tree-structured universal background model

The majority of state-of-the-art speaker recognition systems (SR) utilize speaker models that are derived from an adapted universal background model (UBM) in the form of a Gaussian mixture model (GMM). This is true for GMM supervector systems, joint factor analysis systems, and most recently i-vector systems. In all of these systems, the posterior probabilities and sufficient statistics calculations represent a computational bottleneck in both enrollment and testing. We propose a multi-layered hash system, employing a tree-structured GMM–UBM which uses Runnalls’ Gaussian mixture reduction technique, in order to reduce the number of these calculations. With this tree-structured hash, we can trade-off reduction in computation with a corresponding degradation of equal error rate (EER). As an example, we reduce this computation by a factor of 15× while incurring less than 10% relative degradation of EER (or 0.3% absolute EER) when evaluated with NIST 2010 speaker recognition evaluation (SRE) telephone data.

Event detection in consumer videos using GMM supervectors and SVMs

Abstract In large-scale multimedia event detection, complex target events are extracted from a large set of consumer-generated web videos taken in unconstrained environments. We devised a multimedia event detection method based on Gaussian mixture model (GMM) supervectors and support vector machines. A GMM supervector consists of the parameters of a GMM for the distribution of low-level features extracted from a video clip. A GMM is regarded as an extension of the bag-of-words framework to a probabilistic framework, and thus, it can be expected to be robust against the data insufficiency problem. We also propose a camera motion cancelled feature, which is a spatio-temporal feature robust against camera motions found in consumer-generated web videos. By combining these methods with the existing features, we aim to construct a high-performance event detection system. The effectiveness of our method is evaluated using TRECVID MED task benchmark.

Robust regression fusion of GMM-UBM and GMM-SVM normalized scores using G729 bit-stream for speaker recognition over IP

A novel approach, based on robust regression with normalized score fusion (namely Normalized Scores following Robust Regression Fusion: NSRRF), is proposed for enhancement of speaker recognition over IP networks, which can be used both in Network Speaker Recognition (NSR) and Distributed Speaker Recognition (DSR) systems. In this framework, it is basically assumed that the speech must be encoded by G729 coder in client side, and then, transmitted at a server side, where the ASR systems are located. The Universal Background Gaussian Mixture Model (GMM-UBM) and Gaussian Supervector (GMM-SVM) with normalized scores are used for speaker recognition. In this work, Mel Frequency Cepstral Coefficient (MFCC) and Linear Prediction Cepstral Coefficient (LPCC), both of these features are derived from Line Spectral Pairs (LSP) extracted from G729 bit-stream over IP, constitute the features vectors. Experimental results, conducted with the LIA SpkDet system based on the ALIZE platform3 using ARADIGITS database, have shown in first that the proposed method using features extracted directly from G729 bit-stream reduces significantly the error rate and outperforms the baseline system in ASR over IP based on the resynthesized (reconstructed) speech obtained from the G729 decoder. In addition, the obtained results show that the proposed approach, based on scores normalization following robust regression fusion technique, achieves the best result and outperform the conventional ASR over IP network.

Minimum classification error-based weighted support vector machine kernels for speaker verification

Support vector machines (SVMs) have been proved to be an effective approach to speaker verification. An appropriate selection of the kernel function is a key issue in SVM-based classification. In this letter, a new SVM-based speaker verification method utilizing weighted kernels in the Gaussian mixture model supervector space is proposed. The weighted kernels are derived by using the discriminative training approach, which minimizes speaker verification errors. Experiments performed on the NIST 2008 speaker recognition evaluation task showed that the proposed approach provides substantially improved performance over the baseline kernel-based method.

Reusing Speech Techniques for Video Semantic Indexing [Applications Corner

Many techniques developed in speech research have been successfully employed in other fields, such as automatic video semantic indexing. In this application, a user submits a textual input query for an desired object or a scene to a search system, which returns video shots that include the object or scene. Recently, a new method using Gaussian-mixture model (GMM) supervectors and support vector machines (SVMs) was proven to be very effective. In this method, speech technology such as speaker verification and adaptation techniques play very important roles.

A Fast and Accurate Video Semantic-Indexing System Using Fast MAP Adaptation and GMM Supervectors

We propose a fast maximum a posteriori (MAP) adaptation method for video semantic indexing that uses Gaussian mixture model (GMM) supervectors. In this method, a tree-structured GMM is utilzed to decrease the computational cost, where only the output probabilities of mixture components close to an input sample are precisely calculated. Experimental evaluation on the TRECVID 2010 dataset demonstrates the effectiveness of the proposed method. The calculation time of the MAP adaptation step is reduced by 76.2% compared with that of a conventional method. The total calculation time is reduced by 56.6% while keeping the same level of the accuracy.

Emotional speech recognition based on SVM with GMM supervector

Emotion recognition from speech is an important field of research in human computer interaction. In this letter the framework of Support Vector Machines (SVM) with Gaussian Mixture Model (GMM) supervector is introduced for emotional speech recognition. Because of the importance of variance in reflecting the distribution of speech, the normalized mean vectors potential to exploit the information from the variance are adopted to form the GMM supervector. Comparative experiments from five aspects are conducted to study their corresponding effect to system performance. The experiment results, which indicate that the influence of number of mixtures is strong as well as influence of duration is weak, provide basis for the train set selection of Universal Background Model (UBM).