Search results for: speech feature vector

Authors: Tudor Barbu

Abstract:

We provide a supervised speech-independent voice recognition technique in this paper. In the feature extraction stage we propose a mel-cepstral based approach. Our feature vector classification method uses a special nonlinear metric, derived from the Hausdorff distance for sets, and a minimum mean distance classifier.

Keywords: Text-independent speaker recognition, mel cepstral analysis, speech feature vector, Hausdorff-based metric, supervised classification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1829

Authors: Driss Guerchi, Fatiha Djebbar

Abstract:

In this work we introduce an efficient method to limit the impact of the hiding process on the quality of the cover speech. Vector quantization of the speech spectral information reduces drastically the number of the secret speech parameters to be embedded in the cover signal. Compared to scalar hiding, vector quantization hiding technique provides a stego signal that is indistinguishable from the cover speech. The objective and subjective performance measures reveal that the current hiding technique attracts no suspicion about the presence of the secret message in the stego speech, while being able to recover an intelligible copy of the secret message at the receiver side.

Keywords: Speech steganography, LSF vector quantization, fast Fourier transform

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1515

Authors: H. B. Kekre, Tanuja K. Sarode

Abstract:

Mostly transforms are used for speech data compressions which are lossy algorithms. Such algorithms are tolerable for speech data compression since the loss in quality is not perceived by the human ear. However the vector quantization (VQ) has a potential to give more data compression maintaining the same quality. In this paper we propose speech data compression algorithm using vector quantization technique. We have used VQ algorithms LBG, KPE and FCG. The results table shows computational complexity of these three algorithms. Here we have introduced a new performance parameter Average Fractional Change in Speech Sample (AFCSS). Our FCG algorithm gives far better performance considering mean absolute error, AFCSS and complexity as compared to others.

Keywords: Vector Quantization, Data Compression, Encoding, , Speech coding.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2403

Authors: Ahmed Kamil Hasan Al-Ali, Bouchra Senadji, Ganesh Naik

Abstract:

We propose a system to real environmental noise and channel mismatch for forensic speaker verification systems. This method is based on suppressing various types of real environmental noise by using independent component analysis (ICA) algorithm. The enhanced speech signal is applied to mel frequency cepstral coefficients (MFCC) or MFCC feature warping to extract the essential characteristics of the speech signal. Channel effects are reduced using an intermediate vector (i-vector) and probabilistic linear discriminant analysis (PLDA) approach for classification. The proposed algorithm is evaluated by using an Australian forensic voice comparison database, combined with car, street and home noises from QUT-NOISE at a signal to noise ratio (SNR) ranging from -10 dB to 10 dB. Experimental results indicate that the MFCC feature warping-ICA achieves a reduction in equal error rate about (48.22%, 44.66%, and 50.07%) over using MFCC feature warping when the test speech signals are corrupted with random sessions of street, car, and home noises at -10 dB SNR.

Keywords: Noisy forensic speaker verification, ICA algorithm, MFCC, MFCC feature warping.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 990

Authors: Daniel I. Morariu, Lucian N. Vintan, Volker Tresp

Abstract:

Text categorization is the problem of classifying text documents into a set of predefined classes. After a preprocessing step, the documents are typically represented as large sparse vectors. When training classifiers on large collections of documents, both the time and memory restrictions can be quite prohibitive. This justifies the application of feature selection methods to reduce the dimensionality of the document-representation vector. In this paper, we present three feature selection methods: Information Gain, Support Vector Machine feature selection called (SVM_FS) and Genetic Algorithm with SVM (called GA_SVM). We show that the best results were obtained with GA_SVM method for a relatively small dimension of the feature vector.

Keywords: Feature Selection, Learning with Kernels, Support Vector Machine, Genetic Algorithm, and Classification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1706

Authors: Safdar Tanweer, Abdul Mobin, Afshar Alam

Abstract:

The performance and analysis of speech recognition system is illustrated in this paper. An approach to recognize the English word corresponding to digit (0-9) spoken by 2 different speakers is captured in noise free environment. For feature extraction, speech Mel frequency cepstral coefficients (MFCC) has been used which gives a set of feature vectors from recorded speech samples. Neural network model is used to enhance the recognition performance. Feed forward neural network with back propagation algorithm model is used. However other speech recognition techniques such as HMM, DTW exist. All experiments are carried out on Matlab.

Keywords: Speech Recognition, MFCC, Neural Network, classifier.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3268

Authors: M. Satya Sai Ram, P. Siddaiah, M. Madhavi Latha

Abstract:

This paper investigates the performance of a speech recognizer in an interactive voice response system for various coded speech signals, coded by using a vector quantization technique namely Multi Switched Split Vector Quantization Technique. The process of recognizing the coded output can be used in Voice banking application. The recognition technique used for the recognition of the coded speech signals is the Hidden Markov Model technique. The spectral distortion performance, computational complexity, and memory requirements of Multi Switched Split Vector Quantization Technique and the performance of the speech recognizer at various bit rates have been computed. From results it is found that the speech recognizer is showing better performance at 24 bits/frame and it is found that the percentage of recognition is being varied from 100% to 93.33% for various bit rates.

Keywords: Linear predictive coding, Speech Recognition, Voice banking, Multi Switched Split Vector Quantization, Hidden Markov Model, Linear Predictive Coefficients.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1846

Authors: Ami Kim, Oubong Gwun, Juwhan Song

Abstract:

In this paper, we proposed the distribution of mesh normal vector direction as a feature descriptor of a 3D model. A normal vector shows the entire shape of a model well. The distribution of normal vectors was sampled in proportion to each polygon's area so that the information on the surface with less surface area may be less reflected on composing a feature descriptor in order to enhance retrieval performance. At the analysis result of ANMRR, the enhancement of approx. 12.4%~34.7% compared to the existing method has also been indicated.

Keywords: Interpolated Normal Vector, Feature Descriptor, 3DModel Retrieval.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1474

Authors: Daniel Morariu, Lucian N. Vintan, Volker Tresp

Abstract:

Text categorization is the problem of classifying text documents into a set of predefined classes. After a preprocessing step, the documents are typically represented as large sparse vectors. When training classifiers on large collections of documents, both the time and memory restrictions can be quite prohibitive. This justifies the application of feature selection methods to reduce the dimensionality of the document-representation vector. In this paper, three feature selection methods are evaluated: Random Selection, Information Gain (IG) and Support Vector Machine feature selection (called SVM_FS). We show that the best results were obtained with SVM_FS method for a relatively small dimension of the feature vector. Also we present a novel method to better correlate SVM kernel-s parameters (Polynomial or Gaussian kernel).

Keywords: Feature Selection, Learning with Kernels, SupportVector Machine, and Classification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1829

Authors: Fawaz S. Al-Anzi, Dia AbuZeina

Abstract:

Speech recognition is of an important contribution in promoting new technologies in human computer interaction. Today, there is a growing need to employ speech technology in daily life and business activities. However, speech recognition is a challenging task that requires different stages before obtaining the desired output. Among automatic speech recognition (ASR) components is the feature extraction process, which parameterizes the speech signal to produce the corresponding feature vectors. Feature extraction process aims at approximating the linguistic content that is conveyed by the input speech signal. In speech processing field, there are several methods to extract speech features, however, Mel Frequency Cepstral Coefficients (MFCC) is the popular technique. It has been long observed that the MFCC is dominantly used in the well-known recognizers such as the Carnegie Mellon University (CMU) Sphinx and the Markov Model Toolkit (HTK). Hence, this paper focuses on the MFCC method as the standard choice to identify the different speech segments in order to obtain the language phonemes for further training and decoding steps. Due to MFCC good performance, the previous studies show that the MFCC dominates the Arabic ASR research. In this paper, we demonstrate MFCC as well as the intermediate steps that are performed to get these coefficients using the HTK toolkit.

Keywords: Speech recognition, acoustic features, Mel Frequency Cepstral Coefficients.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1973

Authors: M. Satya Sai Ram, P. Siddaiah, M. Madhavi Latha

Abstract:

Vector quantization is a powerful tool for speech coding applications. This paper deals with LPC Coding of speech signals which uses a new technique called Multi Switched Split Vector Quantization (MSSVQ), which is a hybrid of Multi, switched, split vector quantization techniques. The spectral distortion performance, computational complexity, and memory requirements of MSSVQ are compared to split vector quantization (SVQ), multi stage vector quantization(MSVQ) and switched split vector quantization (SSVQ) techniques. It has been proved from results that MSSVQ has better spectral distortion performance, lower computational complexity and lower memory requirements when compared to all the above mentioned product code vector quantization techniques. Computational complexity is measured in floating point operations (flops), and memory requirements is measured in (floats).

Keywords: Linear predictive Coding, Multi stage vectorquantization, Switched Split vector quantization, Split vectorquantization, Line Spectral Frequencies (LSF).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1672

Authors: M. Satya Sai Ram, P. Siddaiah, M. Madhavi Latha

Abstract:

Vector quantization is a powerful tool for speech coding applications. This paper deals with LPC Coding of speech signals which uses a new technique called Multi Switched Split Vector Quantization, This is a hybrid of two product code vector quantization techniques namely the Multi stage vector quantization technique, and Switched split vector quantization technique,. Multi Switched Split Vector Quantization technique quantizes the linear predictive coefficients in terms of line spectral frequencies. From results it is proved that Multi Switched Split Vector Quantization provides better trade off between bitrate and spectral distortion performance, computational complexity and memory requirements when compared to Switched Split Vector Quantization, Multi stage vector quantization, and Split Vector Quantization techniques. By employing the switching technique at each stage of the vector quantizer the spectral distortion, computational complexity and memory requirements were greatly reduced. Spectral distortion was measured in dB, Computational complexity was measured in floating point operations (flops), and memory requirements was measured in (floats).

Keywords: Unconstrained vector quantization, Linear predictiveCoding, Split vector quantization, Multi stage vector quantization, Switched Split vector quantization, Line Spectral Frequencies.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1741

Authors: Mousmita Sarma, Krishna Dutta, Kandarpa Kumar Sarma

Abstract:

Speech corpus is one of the major components in a Speech Processing System where one of the primary requirements is to recognize an input sample. The quality and details captured in speech corpus directly affects the precision of recognition. The current work proposes a platform for speech corpus generation using an adaptive LMS filter and LPC cepstrum, as a part of an ANN based Speech Recognition System which is exclusively designed to recognize isolated numerals of Assamese language- a major language in the North Eastern part of India. The work focuses on designing an optimal feature extraction block and a few ANN based cooperative architectures so that the performance of the Speech Recognition System can be improved.

Keywords: Filter, Feature, LMS, LPC, Cepstrum, ANN.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2385

Authors: H. F. A. Abdul Ghani, R. R. Porle

Abstract:

Speech to text in Malay language is a system that converts Malay speech into text. The Malay language recognition system is still limited, thus, this paper aims to investigate the performance of ten Malay words obtained from the online Malay news. The methodology consists of three stages, which are preprocessing, feature extraction, and speech classification. In preprocessing stage, the speech samples are filtered using pre emphasis. After that, feature extraction method is applied to the samples using Mel Frequency Cepstrum Coefficient (MFCC). Lastly, speech classification is performed using Feedforward Neural Network (FFNN). The accuracy of the classification is further investigated based on the hidden layer size. From experimentation, the classifier with 40 hidden neurons shows the highest classification rate which is 94%.  

Keywords: Feed-Forward Neural Network, FFNN, Malay speech recognition, Mel Frequency Cepstrum Coefficient, MFCC, speech-to-text.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 746

Authors: Cheima Ben Soltane, Ittansa Yonas Kelbesa

Abstract:

Speaker Identification (SI) is the task of establishing identity of an individual based on his/her voice characteristics. The SI task is typically achieved by two-stage signal processing: training and testing. The training process calculates speaker specific feature parameters from the speech and generates speaker models accordingly. In the testing phase, speech samples from unknown speakers are compared with the models and classified. Even though performance of speaker identification systems has improved due to recent advances in speech processing techniques, there is still need of improvement. In this paper, a Closed-Set Tex-Independent Speaker Identification System (CISI) based on a Multiple Classifier System (MCS) is proposed, using Mel Frequency Cepstrum Coefficient (MFCC) as feature extraction and suitable combination of vector quantization (VQ) and Gaussian Mixture Model (GMM) together with Expectation Maximization algorithm (EM) for speaker modeling. The use of Voice Activity Detector (VAD) with a hybrid approach based on Short Time Energy (STE) and Statistical Modeling of Background Noise in the pre-processing step of the feature extraction yields a better and more robust automatic speaker identification system. Also investigation of Linde-Buzo-Gray (LBG) clustering algorithm for initialization of GMM, for estimating the underlying parameters, in the EM step improved the convergence rate and systems performance. It also uses relative index as confidence measures in case of contradiction in identification process by GMM and VQ as well. Simulation results carried out on voxforge.org speech database using MATLAB highlight the efficacy of the proposed method compared to earlier work.

Keywords: Feature Extraction, Speaker Modeling, Feature Matching, Mel Frequency Cepstrum Coefficient (MFCC), Gaussian mixture model (GMM), Vector Quantization (VQ), Linde-Buzo-Gray (LBG), Expectation Maximization (EM), pre-processing, Voice Activity Detection (VAD), Short Time Energy (STE), Background Noise Statistical Modeling, Closed-Set Tex-Independent Speaker Identification System (CISI).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1887

Authors: Thaweesak Yingthawornsuk, Chusak Thanawattano

Abstract:

Acoustical properties of speech have been shown to be related to mental states of speaker with symptoms: depression and remission. This paper describes way to address the issue of distinguishing depressed patients from remitted subjects based on measureable acoustics change of their spoken sound. The vocal-tract related frequency characteristics of speech samples from female remitted and depressed patients were analyzed via speech processing techniques and consequently, evaluated statistically by cross-validation with Support Vector Machine. Our results comparatively show the classifier's performance with effectively correct separation of 93% determined from testing with the subjectbased feature model and 88% from the frame-based model based on the same speech samples collected from hospital visiting interview sessions between patients and psychiatrists.

Keywords: Depression, SVM, Vocal Extract, Vocal Tract

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1541

Authors: R. Krishnamoorthi, N. Kannan

Abstract:

In this paper, a new algorithm for generating codebook is proposed for vector quantization (VQ) in image coding. The significant features of the training image vectors are extracted by using the proposed Orthogonal Polynomials based transformation. We propose to generate the codebook by partitioning these feature vectors into a binary tree. Each feature vector at a non-terminal node of the binary tree is directed to one of the two descendants by comparing a single feature associated with that node to a threshold. The binary tree codebook is used for encoding and decoding the feature vectors. In the decoding process the feature vectors are subjected to inverse transformation with the help of basis functions of the proposed Orthogonal Polynomials based transformation to get back the approximated input image training vectors. The results of the proposed coding are compared with the VQ using Discrete Cosine Transform (DCT) and Pairwise Nearest Neighbor (PNN) algorithm. The new algorithm results in a considerable reduction in computation time and provides better reconstructed picture quality.

Keywords: Orthogonal Polynomials, Image Coding, Vector Quantization, TSVQ, Binary Tree Classifier

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2149

Authors: Ick Hoon Jang, Hoon Jae Lee, Dae Hoon Kwon, Ui Young Pak

Abstract:

In this paper, we propose a texture feature-based language identification using wavelet-domain BDIP (block difference of inverse probabilities) and BVLC (block variance of local correlation coefficients) features and FFT (fast Fourier transform) feature. In the proposed method, wavelet subbands are first obtained by wavelet transform from a test image and denoised by Donoho-s soft-thresholding. BDIP and BVLC operators are next applied to the wavelet subbands. FFT blocks are also obtained by 2D (twodimensional) FFT from the blocks into which the test image is partitioned. Some significant FFT coefficients in each block are selected and magnitude operator is applied to them. Moments for each subband of BDIP and BVLC and for each magnitude of significant FFT coefficients are then computed and fused into a feature vector. In classification, a stabilized Bayesian classifier, which adopts variance thresholding, searches the training feature vector most similar to the test feature vector. Experimental results show that the proposed method with the three operations yields excellent language identification even with rather low feature dimension.

Keywords: BDIP, BVLC, FFT, language identification, texture feature, wavelet transform.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2149

Authors: Daehwan Kim, Jaemin Kim, Seongwon Cho, Yongsuk Jang, Sun-Tae Chung, Boo-Gyoun Kim

Abstract:

AAM has been successfully applied to face alignment, but its performance is very sensitive to initial values. In case the initial values are a little far distant from the global optimum values, there exists a pretty good possibility that AAM-based face alignment may converge to a local minimum. In this paper, we propose a progressive AAM-based face alignment algorithm which first finds the feature parameter vector fitting the inner facial feature points of the face and later localize the feature points of the whole face using the first information. The proposed progressive AAM-based face alignment algorithm utilizes the fact that the feature points of the inner part of the face are less variant and less affected by the background surrounding the face than those of the outer part (like the chin contour). The proposed algorithm consists of two stages: modeling and relation derivation stage and fitting stage. Modeling and relation derivation stage first needs to construct two AAM models: the inner face AAM model and the whole face AAM model and then derive relation matrix between the inner face AAM parameter vector and the whole face AAM model parameter vector. In the fitting stage, the proposed algorithm aligns face progressively through two phases. In the first phase, the proposed algorithm will find the feature parameter vector fitting the inner facial AAM model into a new input face image, and then in the second phase it localizes the whole facial feature points of the new input face image based on the whole face AAM model using the initial parameter vector estimated from using the inner feature parameter vector obtained in the first phase and the relation matrix obtained in the first stage. Through experiments, it is verified that the proposed progressive AAM-based face alignment algorithm is more robust with respect to pose, illumination, and face background than the conventional basic AAM-based face alignment algorithm.

Keywords: Face Alignment, AAM, facial feature detection, model matching.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1639

Authors: Yong Zhang, Ruimin Hu

Abstract:

This paper presents a vocoder to obtain high quality synthetic speech at 600 bps. To reduce the bit rate, the algorithm is based on a sinusoidally excited linear prediction model which extracts few coding parameters, and three consecutive frames are grouped into a superframe and jointly vector quantization is used to obtain high coding efficiency. The inter-frame redundancy is exploited with distinct quantization schemes for different unvoiced/voiced frame combinations in the superframe. Experimental results show that the quality of the proposed coder is better than that of 2.4kbps LPC10e and achieves approximately the same as that of 2.4kbps MELP and with high robustness.

Keywords: Speech coding, Vector quantization, linear predicition, Mixed sinusoidal excitation

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2188

Authors: Rehab F. Abdel-Kader, Rabab M. Ramadan, Rawya Y. Rizk

Abstract:

The recognition of human faces, especially those with different orientations is a challenging and important problem in image analysis and classification. This paper proposes an effective scheme for rotation invariant face recognition using Log-Polar Transform and Discrete Cosine Transform combined features. The rotation invariant feature extraction for a given face image involves applying the logpolar transform to eliminate the rotation effect and to produce a row shifted log-polar image. The discrete cosine transform is then applied to eliminate the row shift effect and to generate the low-dimensional feature vector. A PSO-based feature selection algorithm is utilized to search the feature vector space for the optimal feature subset. Evolution is driven by a fitness function defined in terms of maximizing the between-class separation (scatter index). Experimental results, based on the ORL face database using testing data sets for images with different orientations; show that the proposed system outperforms other face recognition methods. The overall recognition rate for the rotated test images being 97%, demonstrating that the extracted feature vector is an effective rotation invariant feature set with minimal set of selected features.

Keywords: Discrete Cosine Transform, Face Recognition, Feature Extraction, Log Polar Transform, Particle SwarmOptimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1873

Authors: M. Debyeche, J.P Haton, A. Houacine

Abstract:

The paper presents a complete discrete statistical framework, based on a novel vector quantization (VQ) front-end process. This new VQ approach performs an optimal distribution of VQ codebook components on HMM states. This technique that we named the distributed vector quantization (DVQ) of hidden Markov models, succeeds in unifying acoustic micro-structure and phonetic macro-structure, when the estimation of HMM parameters is performed. The DVQ technique is implemented through two variants. The first variant uses the K-means algorithm (K-means- DVQ) to optimize the VQ, while the second variant exploits the benefits of the classification behavior of neural networks (NN-DVQ) for the same purpose. The proposed variants are compared with the HMM-based baseline system by experiments of specific Arabic consonants recognition. The results show that the distributed vector quantization technique increase the performance of the discrete HMM system.

Keywords: Hidden Markov Model, Vector Quantization, Neural Network, Speech Recognition, Arabic Language

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2056

Authors: Laila Y. Fannas, Ahmed Y. Ben Sasi

Abstract:

This research presents a handwritten signature recognition based on angle feature vector using Artificial Neural Network (ANN). Each signature image will be represented by an Angle vector. The feature vector will constitute the input to the ANN. The collection of signature images will be divided into two sets. One set will be used for training the ANN in a supervised fashion. The other set which is never seen by the ANN will be used for testing. After training, the ANN will be tested for recognition of the signature. When the signature is classified correctly, it is considered correct recognition otherwise it is a failure.

Keywords: Signature Recognition, Artificial Neural Network, Angle Features.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2495

Authors: Nebi Gedik

Abstract:

Regardless of age or gender, cardiovascular illnesses are a serious health concern because of things like poor eating habits, stress, a sedentary lifestyle, hard work schedules, alcohol use, and weight. It tends to happen suddenly and has a high rate of recurrence. Machine learning models can be implemented to assist healthcare systems in the accurate detection and diagnosis of cardiovascular disease (CVD) in patients. Improved heart failure prediction is one of the primary goals of researchers using the heart disease dataset. The purpose of this study is to identify the feature or features that offer the best classification prediction for CVD detection. The support vector machine classifier is used to compare each feature's performance. It has been determined which feature produces the best results.

Keywords: Cardiovascular disease, feature extraction, supervised learning, support vector machine.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 170

Authors: Y. Deng, X. Li, C. Kwan, B. Raj, R. Stern

Abstract:

The current speech interfaces in many military applications may be adequate for native speakers. However, the recognition rate drops quite a lot for non-native speakers (people with foreign accents). This is mainly because the nonnative speakers have large temporal and intra-phoneme variations when they pronounce the same words. This problem is also complicated by the presence of large environmental noise such as tank noise, helicopter noise, etc. In this paper, we proposed a novel continuous acoustic feature adaptation algorithm for on-line accent and environmental adaptation. Implemented by incremental singular value decomposition (SVD), the algorithm captures local acoustic variation and runs in real-time. This feature-based adaptation method is then integrated with conventional model-based maximum likelihood linear regression (MLLR) algorithm. Extensive experiments have been performed on the NATO non-native speech corpus with baseline acoustic model trained on native American English. The proposed feature-based adaptation algorithm improved the average recognition accuracy by 15%, while the MLLR model based adaptation achieved 11% improvement. The corresponding word error rate (WER) reduction was 25.8% and 2.73%, as compared to that without adaptation. The combined adaptation achieved overall recognition accuracy improvement of 29.5%, and WER reduction of 31.8%, as compared to that without adaptation.

Keywords: speaker adaptation; environment adaptation; robust speech recognition; SVD; non-native speech recognition

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3217

Authors: Eric Jiang

Abstract:

As a popular rank-reduced vector space approach, Latent Semantic Indexing (LSI) has been used in information retrieval and other applications. In this paper, an LSI-based content vector model for text classification is presented, which constructs multiple augmented category LSI spaces and classifies text by their content. The model integrates the class discriminative information from the training data and is equipped with several pertinent feature selection and text classification algorithms. The proposed classifier has been applied to email classification and its experiments on a benchmark spam testing corpus (PU1) have shown that the approach represents a competitive alternative to other email classifiers based on the well-known SVM and naïve Bayes algorithms.

Keywords: Feature Selection, Latent Semantic Indexing, Text Classification, Vector Space Model.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1885

Authors: M. Kalamani, S. Valarmathy, M. Krishnamoorthi

Abstract:

In this paper, Fuzzy C-Means clustering with Expectation Maximization-Gaussian Mixture Model based hybrid modeling algorithm is proposed for Continuous Tamil Speech Recognition. The speech sentences from various speakers are used for training and testing phase and objective measures are between the proposed and existing Continuous Speech Recognition algorithms. From the simulated results, it is observed that the proposed algorithm improves the recognition accuracy and F-measure up to 3% as compared to that of the existing algorithms for the speech signal from various speakers. In addition, it reduces the Word Error Rate, Error Rate and Error up to 4% as compared to that of the existing algorithms. In all aspects, the proposed hybrid modeling for Tamil speech recognition provides the significant improvements for speechto- text conversion in various applications.

Keywords: Speech Segmentation, Feature Extraction, Clustering, HMM, EM-GMM, CSR.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2139

Authors: Ahmed Al-Ani

Abstract:

The Ant Colony Optimization (ACO) is a metaheuristic inspired by the behavior of real ants in their search for the shortest paths to food sources. It has recently attracted a lot of attention and has been successfully applied to a number of different optimization problems. Due to the importance of the feature selection problem and the potential of ACO, this paper presents a novel method that utilizes the ACO algorithm to implement a feature subset search procedure. Initial results obtained using the classification of speech segments are very promising.

Keywords: Ant Colony Optimization, ant systems, feature selection, pattern recognition.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3143

Authors: D. Karthikeyan, G. Balakrishnan

Abstract:

Human faces, as important visual signals, express a significant amount of nonverbal info for usage in human-to-human communication. Age, specifically, is more significant among these properties. Human age estimation using facial image analysis as an automated method which has numerous potential real‐world applications. In this paper, an automated age estimation framework is presented. Support Vector Regression (SVR) strategy is utilized to investigate age prediction. This paper depicts a feature extraction taking into account Gray Level Co-occurrence Matrix (GLCM), which can be utilized for robust face recognition framework. It applies GLCM operation to remove the face's features images and Active Appearance Models (AAMs) to assess the human age based on image. A fused feature technique and SVR with GA optimization are proposed to lessen the error in age estimation.

Keywords: Support vector regression, feature extraction, gray level co-occurrence matrix, active appearance models.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1314

Authors: Wernhuar Tarng, Yuan-Yuan Chen, Chien-Lung Li, Kun-Rong Hsie, Mingteh Chen

Abstract:

An emotional speech recognition system for the applications on smart phones was proposed in this study to combine with 3G mobile communications and social networks to provide users and their groups with more interaction and care. This study developed a mechanism using the support vector machines (SVM) to recognize the emotions of speech such as happiness, anger, sadness and normal. The mechanism uses a hierarchical classifier to adjust the weights of acoustic features and divides various parameters into the categories of energy and frequency for training. In this study, 28 commonly used acoustic features including pitch and volume were proposed for training. In addition, a time-frequency parameter obtained by continuous wavelet transforms was also used to identify the accent and intonation in a sentence during the recognition process. The Berlin Database of Emotional Speech was used by dividing the speech into male and female data sets for training. According to the experimental results, the accuracies of male and female test sets were increased by 4.6% and 5.2% respectively after using the time-frequency parameter for classifying happy and angry emotions. For the classification of all emotions, the average accuracy, including male and female data, was 63.5% for the test set and 90.9% for the whole data set.

Keywords: Smart phones, emotional speech recognition, socialnetworks, support vector machines, time-frequency parameter, Mel-scale frequency cepstral coefficients (MFCC).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1842