Search

Under the high-speed condition, the fault diagnosis of rolling bearing is difficult due to parameter limitation and local optimization. To solve these problems, a fault diagnosis method of the whole life cycle based on wavelet thresholding denoising, genetic algorithm-variational mode decomposition (GA-VMD) and improved grey wolf optimizer-least squares support vector machines (IGWO-LSSVM) is proposed. The nonlinear convergence factor is adopted to improve the global search ability of GWO, and then LSSVM is optimized by IGWO, which is used to improve the fault identification accuracy. Based on the high-speed rolling bearing test rig, this method is applied to fault diagnosis of high-speed rolling bearing in the whole life cycle. The kurtosis and approximate entropy indexes are adopted as the basis for the stage division in the degradation process. The fault diagnosis of rolling bearing in the whole life cycle is completed from the five aspects of data acquisition, stage division, data preprocessing, fault feature extraction and fault feature identification. The results show that the maximum identification accuracy of this method can reach 97 %. [ABSTRACT FROM AUTHOR]

The i-vector framework based system is one of the most popular systems in speaker identification (SID). In this system, session compensation is usually employed first and then the classifier. For any session-compensated representation of i-vector, there is a corresponding identification result, so that both the stages are related. However, in current SID systems, session compensation and classifier are usually optimized independently. An incomplete knowledge about the session compensation to the identification task may lead to involving uncertainties. In this paper, we propose a bilevel framework to jointly optimize session compensation and classifier to enhance the relationship between the two stages. In this framework, we use the sparse coding (SC) to obtain the session-compensated feature by learning an overcomplete dictionary, and employ the softmax classifier and support vector machine (SVM) in classifying respectively. Moreover, we present a joint optimization of the dictionary and classifier parameters under a discriminative criterion for classifier with conditions for SC. In addition, the proposed methods are evaluated on the King-ASR-010, VoxCeleb and RSR2015 databases. Compared with typical session compensation techniques, such as linear discriminant analysis (LDA) and nonparametric discriminant analysis (NDA), our methods can be more robust to complex session variability. Moreover, compared with the typical classifiers in i-vector framework, i.e. the cosine distance scoring (CDS) and probabilistic linear discriminant analysis (PLDA), our methods can be more suitable for SID (multiclass task). [ABSTRACT FROM AUTHOR]

Aiming at the problem that weak faults in rolling bearings make effective fault diagnosis difficult under strong noise, this paper proposes a multilevel denoising technology based on improved singular value decomposition (ISVD) and intrinsic timescale decomposition (ITD), combined with an improved deep residual network (ResNet), for fault diagnosis in rolling bearings. Firstly, the difference ratio (DR) index is introduced to optimize singular value decomposition, combined with ITD for multilevel denoising of strong noise signals. Effective fault information in bearing vibration signals is extracted and converted into grayscale images. Secondly, the multi-scale feature extraction module (MFE-Module) is introduced to enhance the feature extraction capability of ResNet, and the support vector machine (SVM) is used instead of the Softmax function to identify and classify the fault features. The experimental results indicate that, compared with other methods, the proposed method can more accurately realize the fault diagnosis of rolling bearings in strong noise environments. • A method is proposed to optimize the singular value decomposition based on the DR index with satisfactory noise reduction effects and a fast processing time. • A multilevel denoising method based on ISVD and ITD is proposed to extract feature information from strong noise signals effectively. • The MResNet-SVM network model is established, and ResNet is optimized by the multi-scale feature extraction module (MFE-Module) and SVM. • The proposed method has excellent noise reduction and fault diagnosis capabilities in a strong noise environment. [ABSTRACT FROM AUTHOR]

The rapidly increasing requirements from context-aware gadgets, like smartphones and intelligent wearable devices, along with applications such as audio archiving, have given a fillip to the research in the field of Acoustic Scene Classification (ASC). The Detection and Classification of Acoustic Scenes and Events (DCASE) challenges have seen systems addressing the problem of ASC from different directions. Some of them could achieve better results than the Mel Frequency Cepstral Coefficients – Gaussian Mixture Model (MFCC-GMM) baseline system. However, a collective decision from all participating systems was found to surpass the accuracy obtained by each system. The simultaneous use of various approaches can exploit the discriminating information in a better way for audio collected from different environments covering audible-frequency range in varying degrees. In this work, we show that the frame-level statistics of some well-known spectral features when fed to Support Vector Machine (SVM) classifier individually, are able to outperform the baseline system of DCASE challenges. Furthermore, we analyzed different methods of combining these features, and also of combining information from two channels when the data is in binaural format. The proposed approach resulted in around 17% and 9% relative improvement in accuracy with respect to the baseline system on the development and evaluation dataset, respectively, from DCASE 2016 ASC task. [ABSTRACT FROM AUTHOR]

A method to improve voicing decision using glottal activity features proposed for statistical parametric speech synthesis. In existing methods, voicing decision relies mostly on fundamental frequency F 0 , which may result in errors when the prediction is inaccurate. Even though F 0 is a glottal activity feature, other features that characterize this activity may help in improving the voicing decision. The glottal activity features used in this work are the strength of excitation (SoE), normalized autocorrelation peak strength (NAPS), and higher-order statistics (HOS). These features obtained from approximated source signals like zero-frequency filtered signal and integrated linear prediction residual. To improve voicing decision and to avoid heuristic threshold for classification, glottal activity features are trained using different statistical learning methods such as the k-nearest neighbor, support vector machine (SVM), and deep belief network. The voicing decision works best with SVM classifier, and its effectiveness is tested using the statistical parametric speech synthesis. The glottal activity features SoE, NAPS, and HOS modeled along with F 0 and Mel-cepstral coefficients in Hidden Markov model and deep neural network to get the voicing decision. The objective and subjective evaluations demonstrate that the proposed method improves the naturalness of synthetic speech. [ABSTRACT FROM AUTHOR]

Nowadays, the popularity of amateur drones (ADr) has increased marginally, and many research areas, such as machine learning, security systems, and the military industry, have been interested in ADrs. The prime goals of this paper are to present a new sound dataset for ADr detection and present an accurate classification model for ADr detection. This research proposes a novel classification model using the proposed skinny pattern and iterative neighborhood component analysis (INCA) feature selector using ADr sounds. The skinny pattern is a new nonlinear pattern that uses the skinny cipher's substitution box (S-Box). Tunable Q-factor wavelet transform (TQWT) and skinny pattern are used together to extract features in multilevel. The most informative features from the generated features are selected using the INCA selector. 16 classifiers were used in five categories to illustrate the general success of the presented model, and these categories are decision tree (DT), discriminant (D), support vector machine (SVM), k nearest neighborhood (kNN), and ensemble classifiers (EC). An ADr sound dataset was collected, and the presented skinny pattern and INCA-based model are applied to the gathered sound dataset to implement the presented model. The best classification accuracy was calculated as 99.72% using Fine kNN. The proposed model was also compared to other ADr detection/classification models, and it outperformed. These obtained results demonstrate the success of the skinny pattern and INCA-based ADr detection model. [ABSTRACT FROM AUTHOR]

Iris recognition systems have demonstrated considerable improvement in recognizing people through their iris patterns. Recent iris recognition systems have focused on images acquired in unconstrained environments. Unconstrained imaging environments allow the capture of iris images at a distance, in motion and under visible wavelength illumination which lead to more noise factors such as off-focus, gaze deviation, and obstruction by eyelids, eyeglasses, hair, lighting and specular reflections. Segmenting irises taken in an unconstrained environment remains a challenging task for iris recognition. In this paper, a new iris segmentation method is developed and tested on UBIRIS.v2 and MICHE iris databases that reflect the challenges in recognition by unconstrained images. This method accurately localizes the iris by a model designed on the basis of the Histograms of Oriented Gradients (HOG) descriptor and Support Vector Machine (SVM), namely HOG-SVM. Based on this localization, iris texture is automatically extracted by means of a cellular automata which evolved via the GrowCut technique. Pre- and post-processing operations are also introduced to ensure higher segmentation accuracy. Extensive experimental results illustrate the effectiveness of the proposed method on unconstrained iris images. [ABSTRACT FROM AUTHOR]

The performance of M -ary quadrature amplitude modulation (QAM) can seriously be degraded by inter-symbol interference (ISI) as the number of levels increases. To mitigate ISI, blind sequence detection (BSD) has very important applications in data transmission systems, particularly where sending a training sequence is disruptive or costly. A new BSD approach of short data in QAM systems using reservoir computing (RC) is presented, together with the detailed theoretical derivation of the algorithm. Its convergence can be guaranteed within a short data packet and, therefore, it works in systems with a much shorter data record and faster time-varying channels. A RC network is constructed to solve the special issue of BSD, with reservoir weight matrix generated via the reduced QR decomposition from the view of receiving signal subspace instead of being selected randomly. The design methods of the activation function and readout function, the variation rule of initial vector which is changed by reservoir weight, and complexity of the proposed algorithm are described, respectively. The readout weight of the RC network is trained and updated by support vector regression (SVR) with a Gaussian insensitive loss function. The correctness and effectiveness of the new approach are verified by simulations, and some special simulation phenomena of the algorithm are discussed. [ABSTRACT FROM AUTHOR]

Following a major disaster, most of the biological characteristics of the victims are typically destroyed, which increases the difficulty of forensic identification. However, palatal rugae cannot be easily destroyed owing to its unique location characteristics, owing to which it has become a new type of biological feature used in forensic identification under special environments. This study proposes a new method of 3D palatal rugae recognition which uses fractional Fourier transform (FRFT). The isometric slicing operation is first used to simplify the complexity of the algorithm and improve the efficiency of the 3D data storage and expression. FRFT is the used to extract the fractional domain features of the profile images obtained by slicing the 3D palatal rugae, and the support vector machine recursive feature elimination (SVM-RFE) method is used for feature selection in the fractional domain. The use of FRFT increases the number of the palatal rugae features extracted from the time-frequency domain, and can improve the effectiveness of feature extraction. The recognition accuracy rate of the proposed 3D palatal rugae recognition system can reach up to 97.8%. The proposed method meets the requirements of 3D palatal rugae recognition, and can help in performing rapid forensic identification. The method proposed in this study is useful in reducing the dependence on human experts and promoting the digitization of a forensic identification system based on palatal rugae. [ABSTRACT FROM AUTHOR]

No-reference (NR)/blind image quality assessment (IQA) metrics play an important role in the area of image processing. Natural scene statistics (NSS) model assumes that natural images possess certain regular statistical properties and is widely used in NR IQA metrics. Most existing NSS-based NR algorithms are achieved by measuring the variation of image statistics, which are characterized by the fitting parameters of NSS model, across different distortions. However, distortions not only change the image statistics, but also disturb the statistical regularity held by natural images. As a result, the distribution of distorted images can not well follow the NSS model. There exists fitting error between the real distribution of the distorted image and the fitted one under certain NSS model. In this paper, the statistical distributions of the distorted images are discussed in detail. We suggest to take the fitting errors into account as well as the fitting parameters for feature extraction, and propose a novel NR IQA algorithm. Experimental results on several image databases demonstrate that the proposed metric performs highly consistent with human visual perception. [ABSTRACT FROM AUTHOR]

The disadvantage of conventional signal processing methods used for Power Quality (PQ) classification is that their mathematical calculations are complex. Therefore, it is necessary to develop simple, effective, and fast signal processing tools to analyze the PQ signal. In this study, a new signal processing method with simple mathematical operations based on the Integer Factor (IF) down sampling/Approximation Derivatives (AD) was developed to analyze PQ signals. IF was employed to analyze the signals at various sampling frequencies, while the AD method was employed to obtain different degrees of detail and approximation coefficients for the signals. In order to prove that the IF/AD signal processing method can perform a fast and detailed signal processing analysis, the IF/AD method and the Discrete Wavelet Transforms (DWT)/Multiresolution method, which is one of the basic signal processing methods, were compared. It was determined that the developed signal processing method has analyzed PQ disturbances in a shorter time and in more detail than DWT/Multiresolution method. After it was determined that the IF/AD method was effective in the analysis of PQ signals, a new classification method based on the IF-AD signal processing approach, the Slime Mould feature selection algorithm, and Support Vector Machine was developed. The classification method was applied to synthetic signals with noise, real-time data generated in the laboratory and PQ online data. The results proved that the developed classification method was successful. Therefore, this study will provide a different approach to PQ classification methods and non-stationary signal analysis studies. [ABSTRACT FROM AUTHOR]

This paper introduces an approach aiming to determine head pose healthiness of computer users. The main contributions of this paper are: 1) Image Quality Assessment (IQA) based Stacked Autoencoder (referred to as IQASAE) which adjusts the value of learning rate based on the quality of images; 2) Head Pose Healthiness Prediction (HPHP) framework which leverages the proposed IQASAE algorithm in combination with image processing operations; 3) A set of features suitable for face analysis applications; 4) Ontology-driven semantic framework which enables further exploiting pose estimation results within applications in synergy with healthcare expert domain knowledge about pose healthiness. Our framework was evaluated on both offline (BIWI and AFLW) and online (our own, collected using Arduino) datasets. Furthermore, it was compared to several state-of-art methods, including Multi-Layer Perceptron (MLP), CART, Random Forest, Convolutional Neural Networks (CNN), Temporal Deep Learning Model (TDLM), hybrid CNN with Support Vector Machine (SVM), Quatnet and Trinet. According to the achieved experimental results, it reaches accuracy up to 79.63% outperforming all of them, except Quatnet and Trinet. However, the main advantages of IQASAE compared to state-of-art methods are: 1) it does not require selection of features, so the processing time is reduced, 2) utilizing angle between chin and mouth reduces training time for SAE, 3) leveraging vector-based feature set to create training data resulted in a significant improvement, especially in offline facial images. [ABSTRACT FROM AUTHOR]

Research on high dimension, low sample size (HDLSS) data has revealed their neighborless nature. This paper addresses the classification of HDLSS image or video data for human activity recognition. Existing approaches often use off-the-shelf classifiers such as nearest neighbor techniques or support vector machines and tend to ignore the geometry of underlying feature distributions. Addressing this issue, we investigate different geometric classifiers and affirm the lack of neighborhoods within HDLSS data. As this undermines proximity based methods and may cause over-fitting for discriminant methods, we propose a QR factorization approach to Nearest Affine Hull (NAH) classification which remedies the HDLSS dilemma and noticeably reduces time and memory requirements of existing methods. We show that the resulting non-parametric models provide smooth decision surfaces and yield efficient and accurate solutions in multiclass HDLSS scenarios. On several action recognition benchmarks, the proposed NAH classifier outperforms other instance based methods and shows competitive or superior performance than SVMs. In addition, for online settings, the proposed NAH method is faster than online SVMs. [ABSTRACT FROM AUTHOR]

In a recent study, we have introduced the problem of identifying cell-phones using recorded speech and shown that speech signals convey information about the source device, making it possible to identify the source with some accuracy. In this paper, we consider recognizing source cell-phone microphones using non-speech segments of recorded speech. Taking an information-theoretic approach, we use Gaussian Mixture Model (GMM) trained with maximum mutual information (MMI) to represent device-specific features. Experimental results using Mel-frequency and linear frequency cepstral coefficients (MFCC and LFCC) show that features extracted from the non-speech segments of speech contain higher mutual information and yield higher recognition rates than those from speech portions or the whole utterance. Identification rate improves from 96.42% to 98.39% and equal error rate (EER) reduces from 1.20% to 0.47% when non-speech parts are used to extract features. Recognition results are provided with classical GMM trained both with maximum likelihood (ML) and maximum mutual information (MMI) criteria, as well as support vector machines (SVMs). Identification under additive noise case is also considered and it is shown that identification rates reduces dramatically in case of additive noise. [ABSTRACT FROM AUTHOR]

Effective fall-detection and classification systems are vital in mitigating severe medical and economical consequences of falls to people in the fall risk groups. One class of such systems is based on wearable sensors. While there is a vast amount of academic work on this class of systems, not much effort has been devoted to the investigation of effective and robust algorithms and like-for-like comparison of state-of-the-art algorithms using a sufficiently large dataset. In this article, fall-direction classification algorithms are presented and compared on an extensive dataset, comprising a total of 1600 fall trials. Eight machine learning classifiers are implemented for fall-direction classification into four basic directions (forward, backward, right, and left). These are, namely, Bayesian decision making (BDM), least squares method (LSM), k -nearest neighbor classifier (k -NN), artificial neural networks (ANNs), support vector machines (SVMs), decision-tree classifier (DTC), random forest (RF), and adaptive boosting or AdaBoost (AB). BDM achieves perfect classification, followed by k -NN, SVM, and RF. Data acquired from only a single motion sensor unit, worn at the waist of the subject, are processed for experimental verification. Four of the classifiers (BDM, LSM, k -NN, and ANN) are modified to handle the presence of data from an unknown class and evaluated on the same dataset. In this robustness analysis, ANN and k -NN yield accuracies above 96.2%. The results obtained in this study are promising in developing real-world fall-classification systems as they enable fast and reliable classification of fall directions. • Employed a fall data set with 1600 simulated fall trials. • Classified falls into four basic directions: forward-backward-right-left. • Implemented and compared accuracy & runtimes of eight machine learning classifiers. • Conducted robustness analysis that includes falls with undefined directions. • Used four of the eight machine learning classifiers and achieved >96% accuracy. [ABSTRACT FROM AUTHOR]

Abstract: This paper proposed various feature extraction procedures to separate crackles and rhonchi of pathological lung sounds from normal lung sounds. The feature extraction process for distinguishing crackles and rhonchus from normal sounds comprises three signal-processing modules with the following functions: (1) was the frequency ratio from the conventional technique of power spectral density (PSD) based on the Welch method. (2) The average instantaneous frequency (IF) and the exchange time of the instantaneous frequency were calculated by the Hilbert Huang transform (HHT). (3) The eigenvalues were obtained from the singular spectrum analysis (SSA) method. In the classification process, a support vector machine (SVM) was used to distinguish the crackles, rhonchus and normal lung sounds. The results showed that the selected features positively represented the characteristic changes in sounds. The PSD frequency ratio and the eigenvalues demonstrate higher classification accuracy (between 90% and 100%) than the calculations of average and exchange time of IF. The calculated features are extremely promising for the evaluation and classification of other biomedical signals as well as other lung sounds. [Copyright &y& Elsevier]

Abstract: The non-negative solution to an underdetermined linear system can be uniquely recovered sometimes, even without imposing any additional sparsity constraints. In this paper, we derive conditions under which a unique non-negative solution for such a system can exist, based on the theory of polytopes. Furthermore, we develop the paradigm of combined sparse representations, where only a part of the coefficient vector is constrained to be non-negative, and the rest is unconstrained (general). We analyze the recovery of the unique, sparsest solution, for combined representations, under three different cases of coefficient support knowledge: (a) the non-zero supports of non-negative and general coefficients are known, (b) the non-zero support of general coefficients alone is known, and (c) both the non-zero supports are unknown. For case (c), we propose the combined orthogonal matching pursuit algorithm for coefficient recovery and derive the deterministic sparsity threshold under which recovery of the unique, sparsest coefficient vector is possible. We quantify the order complexity of the algorithms, and examine their performance in exact and approximate recovery of coefficients under various conditions of noise. Furthermore, we also obtain their empirical phase transition characteristics. We show that the basis pursuit algorithm, with partial non-negative constraints, and the proposed greedy algorithm perform better in recovering the unique sparse representation when compared to their unconstrained counterparts. Finally, we demonstrate the utility of the proposed methods in recovering images corrupted by saturation noise. [Copyright &y& Elsevier]

Abstract: This paper presents a review in the form of a unified framework for tackling estimation problems in Digital Signal Processing (DSP) using Support Vector Machines (SVMs). The paper formalizes our developments in the area of DSP with SVM principles. The use of SVMs for DSP is already mature, and has gained popularity in recent years due to its advantages over other methods: SVMs are flexible non-linear methods that are intrinsically regularized and work well in low-sample-sized and high-dimensional problems. SVMs can be designed to take into account different noise sources in the formulation and to fuse heterogeneous information sources. Nevertheless, the use of SVMs in estimation problems has been traditionally limited to its mere use as a black-box model. Noting such limitations in the literature, we take advantage of several properties of Mercerʼs kernels and functional analysis to develop a family of SVM methods for estimation in DSP. Three types of signal model equations are analyzed. First, when a specific time-signal structure is assumed to model the underlying system that generated the data, the linear signal model (so-called Primal Signal Model formulation) is first stated and analyzed. Then, non-linear versions of the signal structure can be readily developed by following two different approaches. On the one hand, the signal model equation is written in Reproducing Kernel Hilbert Spaces (RKHS) using the well-known RKHS Signal Model formulation, and Mercerʼs kernels are readily used in SVM non-linear algorithms. On the other hand, in the alternative and not so common Dual Signal Model formulation, a signal expansion is made by using an auxiliary signal model equation given by a non-linear regression of each time instant in the observed time series. These building blocks can be used to generate different novel SVM-based methods for problems of signal estimation, and we deal with several of the most important ones in DSP. We illustrate the usefulness of this methodology by defining SVM algorithms for linear and non-linear system identification, spectral analysis, non-uniform interpolation, sparse deconvolution, and array processing. The performance of the developed SVM methods is compared to standard approaches in all these settings. The experimental results illustrate the generality, simplicity, and capabilities of the proposed SVM framework for DSP. [Copyright &y& Elsevier]

Abstract: Detecting and visualizing nonlinear interactive effects of Single Nucleotide Polymorphisms (SNPs) or epistatic interactions are important topics of signal processing having great mathematical and computational challenges. To address these problems, a three-stage method, epiMiner (epistasis Miner), is proposed based on co-information theory. In screening stage, Co-Information Index (CII) is employed to visualize and rank contributions of individual SNPs to the phenotype, with the number of top ranking SNPs retained to next stage specified by users directly or a support vector machine classifier automatically. In testing stage, co-information and co-information based permutation test are conducted sequentially to search epistatic interactions within the retained SNPs, with the results then ranked by their p-values. For further characterizing broader epistasis landscape, a visualizing stage is designed to dynamically construct epistasis networks by linking pairs of the retained SNPs if their co-information values with respect to the phenotype are stronger than thresholds. The performance of epiMiner is compared with existing methods on a diverse range of simulated data sets containing several epistasis models. Results demonstrate that epiMiner is effective in detecting and visualizing epistatic interactions. In addition, the application of epiMiner on a real Age-related Macular Degeneration (AMD) data set provides several new clues for the exploration of causative factors of AMD. The Matlab version of epiMiner software is available free online at https://sourceforge.net/projects/epiminer/files/. [Copyright &y& Elsevier]

In this paper, the intrinsic characteristics of speech modulations are estimated to propose the instant modulation spectral features for efficient emotion recognition. This feature representation is based on single frequency filtering (SFF) technique and higher order nonlinear energy operator. The speech signal is decomposed into frequency sub-bands using SFF, and associated nonlinear energies are estimated with higher order nonlinear energy operator. Then, the feature vector is realized using cepstral analysis. The high-resolution property of SFF technique is exploited to extract the amplitude envelope of the speech signal at a selected frequency with good time-frequency resolution. The fourth order nonlinear energy operator provides noise robustness in estimating the modulation components. The proposed feature set is tested for the emotion recognition task using the i-vector model with the probabilistic linear discriminant scoring scheme, support vector machine and random forest classifiers. The results demonstrate that the performance of this feature representation is better than the widely used spectral and prosody features, achieving detection accuracy of 85.75% , 59.88% , and 65.78% on three emotional databases, EMODB, FAU-AIBO, and IEMOCAP, respectively. Further, the proposed features are found to be robust in the presence of additive white Gaussian and vehicular noises. [ABSTRACT FROM AUTHOR]

Abstract: Recently, finger-vein recognition has received considerable attention. It is widely used in many applications because of its numerous advantages, such as the small capture device, high accuracy, and user convenience. Nevertheless, finger-vein recognition faces a number of challenges. One critical issue is the use of fake finger-vein images to carry out system attacks. To overcome this problem, we propose a new fake finger-vein image-detection method based on the analysis of finger-vein images in both the frequency and spatial domains. This research is novel in five key ways. First, very little research has been conducted to date on fake finger-vein image detection. We construct a variety of fake finger-vein images, printed on A4 paper, matte paper, and overhead projector film, with which we evaluate the performance of our system. Second, because our proposed method is based on a single captured image, rather than a series of successive images, the processing time is short, no additional image alignment is required, and it is very convenient for users. Third, our proposed method is software-based, and can thus be easily implemented in various finger-vein recognition systems without special hardware. Fourth, Fourier transform features in the frequency domain are used for the detection of fake finger-vein images; further, both spatial and frequency characteristics from Haar and Daubechies wavelet transforms are used for fake finger-vein image detection. Fifth, the detection accuracy of fake finger-vein images is enhanced by combining the features of the Fourier transform and Haar and Daubechies wavelet transforms based on support vector machines. Experimental results indicate that the equal error rate of fake finger-vein image detection with our proposed method is lower than that with a Fourier transform, wavelet transform, or other fusion methods. [Copyright &y& Elsevier]

Abstract: Pulmonary crackles are used as indicators for the diagnosis of different pulmonary disorders in auscultation. Crackles are very common adventitious transient sounds. From the characteristics of crackles such as timing and number of occurrences, the type and the severity of the pulmonary diseases may be assessed. In this study, a method is proposed for crackle detection. In this method, various feature sets are extracted using time–frequency and time–scale analysis from pulmonary signals. In order to understand the effect of using different window and wavelet types in time–frequency and time–scale analysis in detecting crackles, different windows and wavelets are tested such as Gaussian, Blackman, Hanning, Hamming, Bartlett, Triangular and Rectangular windows for time–frequency analysis and Morlet, Mexican Hat and Paul wavelets for time–scale analysis. The extracted feature sets, both individually and as an ensemble of networks, are fed into three different machine learning algorithms: Support Vector Machines, k-Nearest Neighbor and Multilayer Perceptron. Moreover, in order to improve the success of the model, prior to the time–frequency/scale analysis, frequency bands containing no-crackle information are removed using dual-tree complex wavelet transform, which is a shift invariant transform with limited redundancy compared to the conventional discrete wavelet transform. The comparative results of individual feature sets and ensemble of sets, which are extracted using different window and wavelet types, for both pre-processed and non-pre-processed data with different machine learning algorithms, are extensively evaluated and compared. [Copyright &y& Elsevier]

Abstract: Audio classification is an important problem in signal processing and pattern recognition with potential applications in audio retrieval, documentation and scene analysis. Common to general signal classification systems, it involves both training and classification (or testing) stages. The performance of an audio classification system, such as its complexity and classification accuracy, depends highly on the choice of the signal features and the classifiers. Several features have been widely exploited in existing methods, such as the mel-frequency cepstrum coefficients (MFCCs), line spectral frequencies (LSF) and short time energy (STM). In this paper, instead of using these well-established features, we explore the potential of sparse features, derived from the dictionary of signal atoms using sparse coding based on e.g. orthogonal matching pursuit (OMP), where the atoms are adapted directly from audio training data using the K-SVD dictionary learning algorithm. To reduce the computational complexity, we propose to perform pooling and sampling operations on the sparse coefficients. Such operations also help to maintain a unified dimension of the signal features, regardless of the various lengths of the training and testing signals. Using the popular support vector machine (SVM) as the classifier, we examine the performance of the proposed classification system for two binary classification problems, namely speech–music classification and male–female speech discrimination and a multi-class problem, speaker identification. The experimental results show that the sparse (max-pooled and average-pooled) coefficients perform better than the classical MFCCs features, in particular, for noisy audio data. [Copyright &y& Elsevier]

Abstract: Dysfluency and stuttering are a break or interruption of normal speech such as repetition, prolongation, interjection of syllables, sounds, words or phrases and involuntary silent pauses or blocks in communication. Stuttering assessment through manual classification of speech dysfluencies is subjective, inconsistent, time consuming and prone to error. This paper proposes an objective evaluation of speech dysfluencies based on the wavelet packet transform with sample entropy features. Dysfluent speech signals are decomposed into six levels by using wavelet packet transform. Sample entropy (SampEn) features are extracted at every level of decomposition and they are used as features to characterize the speech dysfluencies (stuttered events). Three different classifiers such as k-nearest neighbor (kNN), linear discriminant analysis (LDA) based classifier and support vector machine (SVM) are used to investigate the performance of the sample entropy features for the classification of speech dysfluencies. 10-fold cross validation method is used for testing the reliability of the classifier results. The effect of different wavelet families on the classification performance is also performed. Experimental results demonstrate that the proposed features and classification algorithms give very promising classification accuracy of 96.67% with the standard deviation of 0.37 and also that the proposed method can be used to help speech language pathologist in classifying speech dysfluencies. [Copyright &y& Elsevier]

Abstract: This paper presents a new method for differential diagnosis of erythemato-squamous diseases based on Genetic Algorithm (GA) wrapped Bayesian Network (BN) Feature Selection (FS). With this aim, a GA based FS algorithm combined in parallel with a BN classifier is proposed. Basically, erythemato-squamous dataset contains six dermatological diseases defined with 34 features. In GA–BN algorithm, GA makes a heuristic search to find most relevant feature model that increase accuracy of BN algorithm with the use of a 10-fold cross-validation strategy. The subsets of features are sequentially used to identify six dermatological diseases via a BN fitting the corresponding data. The algorithm, in this case, produces 99.20% classification accuracy in the diagnosis of erythemato-squamous diseases. The strength of feature model generated for BN is furthermore tested with the use of Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), Simple Logistics (SL) and Functional Decision Tree (FT). The resultant classification accuracies of algorithms are 98.36%, 97.00%, 98.36% and 97.81% respectively. On the other hand, BN algorithm with classification accuracy of 99.20% is quite a high diagnosis performance for erythemato-squamous diseases. The proposed algorithm makes no more than 3 misclassifications out of 366 instances. Furthermore, FS power of GA is also compared with two alternative search algorithms, i.e. Best First (BF) and Sequential Floating (SF). The obtained results have all together shown that the proposed GA–BN based FS and prediction strategy is very promising in diagnosis of erythemato-squamous diseases. [Copyright &y& Elsevier]

Abstract: To solve the speaker independent emotion recognition problem, a three-level speech emotion recognition model is proposed to classify six speech emotions, including sadness, anger, surprise, fear, happiness and disgust from coarse to fine. For each level, appropriate features are selected from 288 candidates by using Fisher rate which is also regarded as input parameter for Support Vector Machine (SVM). In order to evaluate the proposed system, principal component analysis (PCA) for dimension reduction and artificial neural network (ANN) for classification are adopted to design four comparative experiments, including Fisher+SVM, PCA+SVM, Fisher+ANN, PCA+ANN. The experimental results proved that Fisher is better than PCA for dimension reduction, and SVM is more expansible than ANN for speaker independent speech emotion recognition. The average recognition rates for each level are 86.5%, 68.5% and 50.2% respectively. [Copyright &y& Elsevier]

Abstract: This paper proposes a new fuzzy classifier (FC)-based face localization approach. The FC used is a self-organizing TS-type fuzzy network with support vector learning (SOTFN-SV). The SOTFN-SV learns consequent parameters using a linear support vector machine to improve generalization ability. The FC is first applied to segment human skin pixels in scaled hue and saturation (hS) color space, after which connected skin-color regions are regarded as face candidates. The FC is then applied to detect and localize faces from the candidates. The proposed FC-based face localization approach uses shape and wavelet-localized focus color features. A best fitting ellipse of each face candidate is found to obtain shape features. Focus color features are extracted from four focus regions, including the two eyes, the mouth, and the face skin-color region. To find these focus color regions, the Haar-wavelet transformation is first applied to the face candidates in the YCb color space to localize all possible pairs of eye candidates. The mouth region is then localized according to its geometric relationship with the eyes. The hS color features of the located eyes, mouth, and face skin are extracted. These focus color features, together with shape features, serve as inputs to another FC for final face localization. Comparisons with various classifiers and face detection methods demonstrate the advantage of the FC-based skin color segmentation and face localization method. [Copyright &y& Elsevier]

Abstract: Taking into account two types of tremor, namely essential tremor (ET) and Parkinsonʼs disease (PD), which are often misdiagnosed in clinical practice, a novel approach using singular value decomposition (SVD) to extract the features of intrinsic mode functions (IMFs) and support vector machine (SVM) is proposed to distinguish between them. The hand acceleration signals of 25 voluntary subjects with two conditions were collected and preprocessed. The empirical mode decomposition (EMD) method decomposed the signals into a number of stationary IMFs. The singular value features of the preceding four IMFs were extracted, and the features were inputted to the SVM classifier, which can recognize the two types of tremor. To comparing, we also used the singular value features of discrete wavelet transform (DWT) as input to the SVM. Cross-validation testing results indicated that the accuracy, sensitivity, and specificify of EMD-SVD features extracted are all remarkable higher than that of DWT-SVD method. Due to the accuracy, sensitivity, and specificify could arrive at 98%, 97.5% and 98.33% respectively, thus, practical guiding significance for diagnosing tremor types in clinic is provided. [Copyright &y& Elsevier]

Abstract: Automatic recognition of communication signal type plays an important role in various applications. Most of the existing recognizers can only identify a few types of communication signal. This paper presents a novel intelligent technique that identifies a variety of digital signal types. Here, a hierarchical support vector machine based structure is proposed as the multiclass classifier. A proper set of the higher order moments (up to eighth) and higher order cumulants (up to eighth) are proposed as the effective features for recognizing of the digital communication signal. A genetic algorithm is used for selecting the suitable parameters of support vector machines. This idea improves the performance of the recognizer, efficiently. Simulation results show that the proposed recognizer has a high success rate for recognition of the different modulations even at very low SNRs. [Copyright &y& Elsevier]

Abstract: In this paper, we present a family of complex-valued support vector classifiers (CSVCs) based on the definition of the complex sign inspired by the modulation in digital communications and the complex-valued kernel functions. We also propose a theorem to construct the complex-valued Mercer kernels and the corresponding kernel function groups. CSVC algorithms include binary (2-state) CSVC (BCSVC), quadrature (4-state) CSVC (QCSVC) and some multi-state CSVCs. In this paper, we focus on QCSVC. For a quadrature complex-valued classification problem, QCSVC is identical to the 4-quadrature amplitude modulation demodulation methods in digital communications. Finally, the simulated experiments confirm the validity and the efficiency of CSVCs. [Copyright &y& Elsevier]

Abstract: Breast cancer can be treated most effectively when detected in its early stage. Due to the superiority to mammography in its ability to detect focal abnormalities in the dense breasts of adolescent women, sonography has become an important adjunct to mammography in breast cancer detection and has been especially useful in distinguishing cysts from solid tumors. In this paper, we develop a novel CAD system based on fuzzy support vector machine to automatically detect and classify mass using ultrasound (US) images. The experimental results show that the proposed system greatly improves the five objective measurements and the area () under the ROC curve compared with those of other classification methods, and radiologist assessments, and the proposed approach will be very valuable for breast cancer control. [Copyright &y& Elsevier]

Probabilistic linear discriminant analysis (PLDA) has achieved good performance in face recognition and speaker recognition. However, the computation of PLDA using the original formulation is inefficient when there are many training data, especially when the dimensionality of the data is high. Faced with this inefficiency issue, we propose scalable formulations for PLDA. The computation of PLDA using the scalable formulations is more efficient than using the original formulation when dealing with many training data. Using the scalable formulations, the PLDA model can significantly outperform other popular classifiers for speaker recognition, such as Support Vector Machine (SVM) and Gaussian Mixture Model (GMM). Besides of directly using PLDA as a classifier, we may also use PLDA as a feature transformation technique. This PLDA-based feature transformation technique can reduce or expand the original feature dimensionality, and at the same time keep the transformed feature vector approximately following the Gaussian distribution. Our experimental results on speaker recognition and acoustic scene classification demonstrate the effectiveness of applying PLDA for feature transformation. It is then promising to combine PLDA with other classification models for improved performance, extending the utility of PLDA to a wider range of areas. [ABSTRACT FROM AUTHOR]

The fault diagnosis of the common rail injector is an important means to ensure the safe operation of the diesel engine. In order to quickly and accurately identify the fault status of common rail injectors, this paper proposes an intelligent fault diagnosis method for common rail injectors based on Composite Hierarchical Dispersion Entropy (CHDE) and Improved Grasshopper Optimization Algorithm based Least Squares Support Vector Machine (IGOA-LSSVM). First, in order to avoid the inherent shortcomings of Hierarchical Dispersion Entropy, we calculate CHDE as a characteristic parameter to construct a fault characteristic set. Then, this paper proposes the IGOA-LSSVM multi-classifier for pattern recognition, which has higher recognition accuracy and stability than other classifiers. Finally, we use the proposed method to analyze the common rail injector failure data. The results show that the proposed method can not only effectively realize the common rail injector intelligent fault diagnosis but also has a higher fault recognition rate than existing methods. [ABSTRACT FROM AUTHOR]

Early fault diagnosis of common rail injectors is essential to reduce diesel engine testing and maintenance costs. Therefore, this paper proposes a new common rail injector early fault diagnosis method, which combines the Maximum Second-order Cyclostationary Blind Deconvolution (CYCBD) optimized by the Seagull Optimization Algorithm (SOA) and Hierarchical Fluctuation Dispersion Entropy (HFDE). First, we use SOA adaptively to seek the optimal filter length of CYCBD and use the optimal CYCBD to filter the fuel pressure signal of the high-pressure fuel pipe. Then, in order to make up for the shortcomings of Multi-scale Fluctuation Dispersion Entropy (MFDE) ignoring high-frequency component information, this paper proposes HFDE to extract the fault characteristics after filtering. Finally, we input the fault characteristics into Least Squares Support Vector Machines (LSSVM) for classification and recognition. Through the analysis of experimental data, the method proposed in this paper can effectively identify the early failure state of the common rail injector. Compared with the existing methods, the proposed method has a higher fault recognition rate. [ABSTRACT FROM AUTHOR]

The heartbeat interval (HBI) signal (RR-time series), and electrocardiogram (ECG) derived respiration (EDR) signal quantify the information about the cardiopulmonary activity, and monitoring these two signals simultaneously will provide more information for the sleep apnea detection. This paper proposes a novel approach to detect sleep apnea using both HBI and EDR signals. The approach consists of the decomposition of both HBI and EDR signals into reconstructed components (RCs) or modes using a data-driven signal processing approach namely, the sliding mode singular spectrum analysis (SM-SSA), extraction of features from each RC, and the use of classifier for the detection of sleep apnea. The features such as the mean and the standard deviation values are extracted from the instantaneous amplitude (IA) and instantaneous frequency (IF) of each RC of both HBI and EDR signals. The classifiers, such as the stacked autoencoder based deep neural network (SAE-DNN), and support vector machine (SVM) are considered to classify normal and apnea episodes using the statistical features obtained from the RCs of HBI and EDR signals. The proposed approach is evaluated using different public databases such as apnea-ECG database, University College Dublin (UCD) database, and Physionet challenge database, respectively. The results demonstrate that the combination of the statistical features and SVM classifier has the sensitivity and specificity values of 82.45% and 79.72%, respectively using the 10-fold cross-validation based selection of training and test instances from the apnea-ECG database. Moreover, for subject-specific cross-validation, the proposed method has an average sensitivity and specificity values of 62.87%, and 81.53%, respectively. The proposed method has achieved the accuracy values of 94.3%, and 72% for per-recording based classification of sleep apnea and normal classes using signals from apnea-ECG and UCD databases. [ABSTRACT FROM AUTHOR]

• MFCC and HFCC features are extracted using mel and human factor (HF) filter banks. • Mel filter bank uses mel frequency scale, while HF filter bank uses mel and ERB. • Mel and HF filter banks have magnitude responses which are triangular in shape. • Likewise, two new filter banks are realized using bark and ERB frequency scales. • The proposed filter banks are used to extract two new cepstral feature variants. Speech emotion recognition (SER) refers to the process of recognizing the emotional state of the speaker from the speech utterance. In earlier studies, wide varieties of cepstral features have been proposed to develop SER systems. The mel-frequency cepstral coefficients (MFCC), and human-factor cepstral coefficients (HFCC) are two popularly used variants of cepstral features. MFCC and HFCC features are extracted from speech signals using mel and human-factor filter banks, respectively. The magnitude response of individual filters in these filter banks is triangular in shape. As a result, these filter banks are referred to as triangular filter banks (TFB) and the corresponding extracted cepstral coefficients can be denoted as TFBCC-M (in case of MFCC) and TFBCC-HF (in case of HFCC). The mel-filter bank (TFB-M) is constructed using mel-scale, while the human-factor filter bank (TFB-HF) is constructed using human factor scale, which is a combination of mel and equivalent rectangular bandwidth (ERB) scales. Similarly, different frequency scales can be used to realize different TFBs to extract different types of TFBCC features. In this direction, this paper proposes two new TFBs denoted as TFB-B and TFB-E, realized using bark and ERB scales to extract new cepstral features referred to as TFBCC-B and TFBCC-E, respectively. The mathematical background to construct the proposed TFB-B and TFB-E is presented. The proposed filter banks are used along with the conventional TFB-M and TFB-HF to extract four different types of TFBCC features. These features are extracted from the emotional speech signals of two databases, namely Berlin database of emotional speech (Emo-DB) and Surrey audio-visual expressed emotion speech database (SAVEE). The extracted features are used to develop speaker-dependent (SD) and speaker-independent (SI) based SER systems using support vector machines. The performance of the respective features is analyzed in terms of isolated and combined usage. The experimental results show that the cepstral features extracted using the proposed TFBs are effective in characterizing and recognizing emotions similar to conventional MFCC and HFCC features. Moreover, the combined use of different cepstral features have resulted to improve the overall recognition performance of SER systems. In case of Emo-DB database, isolated use of the proposed TFBCC-B and TFBCC-E features achieve recognition accuracies of 83.23% and 81.99% for SD scenario, and 75% and 60.94% for SI scenario, respectively. Similarly, for SAVEE database, the recognition accuracies of 75% and 66.67% for SD scenario, and 44.17% and 55% for SI scenario are achieved. In case of Emo-DB database, the maximum recognition accuracies of 86.96% (for different combinations of conventional and proposed features namely, TFBCC-{(M+E), (M+B+E), (HF+B+E), (M+HF+B+E)}) and 77.08% (for combination TFBCC-(M+B+E)) are achieved for SD and SI scenarios, respectively. Similarly, for SAVEE database, the maximum recognition accuracies of 77.08% (for combination TFBCC-(M+HF+E)), and 55.83% (for combination TFBCC-(B+E)) are achieved for SD and SI scenarios, respectively. [ABSTRACT FROM AUTHOR]

In this paper, we study anomaly detection in a distributed network of nodes and introduce a novel algorithm based on Support Vector Machines (SVMs). We first reformulate the conventional SVM optimization problem for a distributed network of nodes. We then directly train the parameters of this SVM architecture in its primal form using a gradient based algorithm in a fully distributed manner, i.e., each node in our network is allowed to communicate only with its neighboring nodes in order to train the parameters. Therefore, we not only obtain a high performing anomaly detection algorithm thanks to strong modeling capabilities of SVMs, but also achieve significantly reduced communication load and computational complexity due to our fully distributed and efficient gradient based training. Here, we provide a training algorithm in a supervised framework, however, we also provide the extensions of our implementation to an unsupervised framework. We illustrate the performance gains achieved by our algorithm via several benchmark real life and synthetic experiments. [ABSTRACT FROM AUTHOR]