Search

In this work, we conducted an empirical comparative study of the performance of text-independent speaker verification in emotional and stressful environments. This work combined deep models with shallow architecture, which resulted in novel hybrid classifiers. Four distinct hybrid models were utilized: deep neural network-hidden Markov model (DNN-HMM), deep neural network-Gaussian mixture model (DNN-GMM), Gaussian mixture model-deep neural network (GMM-DNN), and hidden Markov model-deep neural network (HMM-DNN). All models were based on novel implemented architecture. The comparative study used three distinct speech datasets: a private Arabic dataset and two public English databases, namely, Speech Under Simulated and Actual Stress (SUSAS) and Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS). The test results of the aforementioned hybrid models demonstrated that the proposed HMM-DNN leveraged the verification performance in emotional and stressful environments. Results also showed that HMM-DNN outperformed all other hybrid models in terms of equal error rate (EER) and area under the curve (AUC) evaluation metrics. The average resulting verification system based on the three datasets yielded EERs of 7.19%, 16.85%, 11.51%, and 11.90% based on HMM-DNN, DNN-HMM, DNN-GMM, and GMM-DNN, respectively. Furthermore, we found that the DNN-GMM model demonstrated the least computational complexity compared to all other hybrid models in both talking environments. Conversely, the HMM-DNN model required the greatest amount of training time. Findings also demonstrated that EER and AUC values depended on the database when comparing average emotional and stressful performances., Comment: 23 pages, 13 figures

This paper proposes a novel idea to governing wind power conversion plants supplying DC loads characterizing an isolated site based on self-excited squirrel-cage induction generators (IG). In this wind power converting application, the induction generator produces an active power from the mechanical power provided by a wind-turbine to variable DC loads through a static converter with an output capacitor under constant voltage levels. For this reason, A specific vector control technique has been developed for controlling the induction machine in an analogous manner with a separated DC machine case. Thus, in order to satisfy the active power demand characterizing a variable DC load at a given rotor mechanical speed, the corresponding control laws are performed in steady-state conditions from a new control variable introduction defined by the ratio of the desired output DC-bus voltage square value and the rotor velocity. Computer simulations validated by experimental results demonstrate that the projected control approach including just one conventional controller ensures excellent tracking performances of the DC-bus voltage to its reference trajectory under simultaneous variations of the load-power demand and the rotor velocity profiles.

Purpose. The aim of this paper is to perform a simple and robust control method based on the well-known sliding control approach for a self-excited induction generator supplying an isolated DC load; this adopted technique does not require much computation and could be easily implemented in practice. In this context, the present work will begin with a mathematical development of this control technique and its application to the self-excited induction generator case. For this purpose, the machine provides the produced active power to the load through a static PWM converter equipped with a single capacitor on the DC side. In order to insure the output DC-bus voltage regulation with respect to the load-power demands and the rotor speed fluctuations, the required stator currents references are computed by considering the reactive power required for the machine core magnetization, the induced voltages through the stator windings and the active power set value obtained from the corresponding sliding mode DC-bus voltage controller. Regarding the nonlinearity of the DC-bus voltage mathematical model and the discontinuity characterizing the converter-machine behavior association, the sliding mode strategy will constitute a perfect tool to sizing the controller structure with high control performances. Results of simulation carried out to demonstrate the proposed control validity are presented., Целью данной статьи является разработка простого и надежного метода управления, основанного на хорошо известном подходе к управлению скольжением для асинхронного генератора с самовозбуждением, питающего изолированную нагрузку постоянного тока; данный принятый метод не требует больших объемов вычислений и может быть легко реализован на практике. В этом контексте данная работа начинается с развития математических основ этого метода управления и его применения в случае асинхронного генератора с самовозбуждением. Для этого машина подает произведенную активную мощность в нагрузку через статический ШИМ-преобразователь, оснащенный единственным конденсатором на стороне постоянного тока. Чтобы обеспечить регулирование выходного напряжения шины постоянного тока с учетом требований к нагрузке и колебаниям скорости вращения ротора, требуемые токи статора рассчитываются с учетом реактивной мощности, необходимой для намагничивания сердечника машины, наведенных напряжений в обмотках статора и заданного значения активной мощности, полученного из соответствующего контроллера напряжения шины постоянного тока в режиме скольжения. Что касается нелинейности математической модели напряжения шины постоянного тока и неоднородности, характеризующей поведение системы «преобразователь-машина», стратегия скользящего режима будет представлять собой идеальный инструмент для определения размеров конструкции контроллера с высокими характеристиками управления. Для демонстрации обоснованности предлагаемого метода контроля, приведены результаты выполненного моделирования.

Robot table tennis systems require a vision system that can track the ball position with low latency and high sampling rate. Altering the ball to simplify the tracking using, for instance, infrared coating changes the physics of the ball trajectory. As a result, table tennis systems use custom tracking systems to track the ball based on heuristic algorithms respecting the real-time constrains applied to RGB images captured with a set of cameras. However, these heuristic algorithms often report erroneous ball positions, and the table tennis policies typically need to incorporate additional heuristics to detect and possibly correct outliers. In this paper, we propose a vision system for object detection and tracking that focuses on reliability while providing real-time performance. Our assumption is that by using multiple cameras, we can find and discard the errors obtained in the object detection phase by checking for consistency with the positions reported by other cameras. We provide an open source implementation of the proposed tracking system to simplify future research in robot table tennis or related tracking applications with strong real-time requirements. We evaluate the proposed system thoroughly in simulation and in the real system, outperforming previous work. Furthermore, we show that the accuracy and robustness of the proposed system increases as more cameras are added. Finally, we evaluate the table tennis playing performance of an existing method in the real robot using the proposed vision system. We measure a slight increase in performance compared to a previous vision system even after removing all the heuristics previously present to filter out erroneous ball observations.

Voltage sag, which is associated to a transitory drop in the root mean square voltage characterizing an electrical source network. During these perturbations, the corresponding electronic customers and devices will suffer from serious operating troubles causing dangerous damages. Purpose. In order to attenuate this disturbance effects, the Controlled Dynamic Voltage Restorer constitutes a very interesting solution among many others that have been proposed. The novelty of the proposed work consists in presenting an enhanced algorithm to control efficiently the dynamic voltage restorer when voltage sag is suddenly occurred. Methods. The proposed algorithm is based on an instantaneous phase locked loop using a multi variable filter to synthesize unitary signals involved in compensation voltages computation relative to the sag apparition. Practical value. A detailed study concerning typical voltage sag, which is consolidated by simulation and experimental results, is conducted to show the used algorithm’s effectiveness to cancel the corresponding voltage sag. Провал напруги, який пов’язаний із тимчасовим падінням середньоквадратичної напруги, характеризує мережу джерел електричної енергії. Під час цих збурень відповідні споживачі (електронні прилади та інші пристрої) будуть страждати від серйозних проблем у їхній експлуатації, що спричиняють небезпечні пошкодження. Призначення. Для того, щоб послабити вплив цих збурень, контрольований динамічний відновник напруги видається дуже цікавим рішенням серед багатьох інших, які були запропоновані. Новизна запропонованої роботи полягає у представленні вдосконаленого алгоритму ефективного управління динамічним відновником напруги, коли раптово відбувається провал напруги. Методи. Запропонований алгоритм базується на миттєвому фазовому замкненому контурі з використанням багатоваріантного фільтра для синтезу унітарних сигналів, що беруть участь у обчисленні напруги компенсації стосовно прояву провалу. Практичне значення. Детальне дослідження стосовно типового провалу напруги, яке узагальнено за допомогою моделювання та експериментальних результатів, проведено, щоб показати ефективність використовуваного алгоритму для ліквідації відповідного провалу напруги.

This work aims at intensifying text-independent speaker identification performance in real application situations such as noisy and emotional talking conditions. This is achieved by incorporating two different modules: a Computational Auditory Scene Analysis CASA based pre-processing module for noise reduction and cascaded Gaussian Mixture Model Convolutional Neural Network GMM-CNN classifier for speaker identification followed by emotion recognition. This research proposes and evaluates a novel algorithm to improve the accuracy of speaker identification in emotional and highly-noise susceptible conditions. Experiments demonstrate that the proposed model yields promising results in comparison with other classifiers when Speech Under Simulated and Actual Stress SUSAS database, Emirati Speech Database ESD, the Ryerson Audio-Visual Database of Emotional Speech and Song RAVDESS database and the Fluent Speech Commands database are used in a noisy environment., Comment: Published in Applied Soft Computing journal

One-class classifier (OCC) is involved for solving different kinds of problems due to its ability to represent a class distribution regardless the remaining classes. Its main advantage for multi-class classification is offering an open system and therefore allows easily extending new classes without retraining OCCs. So far, hidden Markov models, support vector machines and neural networks are the most used classifiers for Arabic word recognition, which provides a system with closed lexicon. In this paper, the OCCs are explored in order to perform an Arabic word recognition system with an open lexicon. Generally, pattern recognition systems designed by a single system suffer from limitations such as the lack of uniqueness and non-universality. Thus, combining multiple systems becomes an attractive research topic for performance and robustness enhancement. Fixed rules are commonly used us combiners for the hybrid OCC ensembles. The present paper aims to propose a combination scheme of OCCs based on the use of fuzzy integral (FI) operators. Furthermore, an alternative framework is proposed to design a parameter-independent and open-lexicon handwritten Arabic word recognition system as well as a new density measure function. Experimental results conducted on Arabic handwritten dataset using different types of OCCs with large number of classes highlight the superiority of FI for hybrid OCC ensembles.

The aim of this paper is to present a study of the efficiency of the electrical part of a wind generation system. Two back-to-back PWM voltage-fed inverters connected between the stator and the rotor are used to allow bidirectional power flow. The second inverter grid side, has a role of a power active filter, to eliminate the harmonic generated by the non linear load, in the same time gives an active and reactive power needed by the rotor of DFIG. The harmonics of switching frequency in the current stator, pose a major problem in the moment where commutations in the diode bridge, to solve this problem, we introduce a small-sized passive LC filter for the purpose of eliminating high-frequency shaft voltage and grid current from a DFIG driven by a voltage-source pulse width-modulation rotor inverter controlled with SVM. The control theory is discussed, and the controller implementation is described. Design criteria are also given. The results of simulation tests show excellent static and dynamic performances.

One-class classifier (OCC) is involved for solving different kinds of problems due to its ability to represent a class distribution regardless the remaining classes. Its main advantage for multi-class classification is offering an open system and therefore allows easily extending new classes without retraining OCCs. So far, hidden Markov models, support vector machines and neural networks are the most used classifiers for Arabic word recognition, which provides a system with closed lexicon. In this paper, the OCCs are explored in order to perform an Arabic word recognition system with an open lexicon. Generally, pattern recognition systems designed by a single system suffer from limitations such as the lack of uniqueness and non-universality.

Graphical abstractDisplay Omitted Currently, writer's soft-biometrics prediction is gaining an important role in various domains related to forensics and anonymous writing identification. The purpose of this work is to develop a robust prediction of the writer's gender, age range and handedness. First, three prediction systems using SVM classifier and different features, that are pixel density, pixel distribution and gradient local binary patterns, are proposed. Since each system performs differently to the others, a combination method that aggregates a robust prediction from individual systems, is proposed. This combination uses Fuzzy MIN and MAX rules to combine membership degrees derived from predictor outputs according to their performances, which are modeled by Fuzzy measures. Experiments are conducted on two Arabic and English public handwriting datasets. The comparison of individual predictors with the state of the art highlights the relevance of proposed features. Besides, the proposed Fuzzy MIN-MAX combination comfortably outperforms individual systems and classical combination rules. Relatively to Sugeno's Fuzzy Integral, it has similar computational complexity while performing better in most cases.

A new system for Handwritten Signature Verification is proposed.The Artificial Immune Recognition System is proposed to achieve the verification task.Two new features that are Gradient Local Binary Patterns and Longest Run Features are proposed.Results obtained on CEDAR and GPDS-100 corpuses reveal the effectiveness of the proposed methods. Natural Immune System offers many interesting features that inspired the design of Artificial Immune Systems (AIS) used to solve various problems of engineering and artificial intelligence. AIS are particularly successful in fault detection and diagnosis applications where anomalies such as errors and failures are assimilated to viruses that should be detected. Thereby, AIS seem suitable to automatically detect forgeries in signature verification systems. This paper proposes a novel method for off-line signature verification that is based on the Artificial Immune Recognition System (AIRS). For feature generation, two different descriptors are proposed to generate signature traits. The first is the Gradient Local Binary Patterns that estimates gradient features based on the LBP neighborhood. The second descriptor is the Longest Run Feature, which describes the signature topology by considering longest suites of text pixels. Performance evaluation is carried out on CEDAR and GPDS-100 datasets. The results obtained showed that the proposed system has promising performance and often comfortably outperforms the state of the art.

Retrieval based on query images supports interesting applications in handwritten document analysis, such as checking manuscripts originality, and authorship. In this respect, writer retrieval systems aim to automatically find all manuscripts belonging to the same author. Presently, we propose a new combination scheme for multiple writer retrieval systems that employ different features and dissimilarities. The proposed combination is founded on writer-independent, SVM dissimilarity learning. For experimental evaluation, three individual systems are proposed each of which, has its specific features. To develop the first system, we propose the Multiscale Histogram Of Templates (M-HOT). For the second system, we introduce the so-called Multi-Gradient Elongated Quinary Pattern (MG-EQP) as new descriptor for handwriting characterization. The third system uses the well-known Run Length Features. Retrieval tests are performed on CVL, ICDAR-2011, ICDAR-2013 and ICDAR-2017 datasets. Furthermore, to highlight the language-independence aspect, experiments are performed on KHATT dataset that contains Arabic handwritten documents. Results obtained evince the effectiveness of the proposed features as well as the combination scheme, which outperforms both individual systems and the state of the art.

Verifying the authenticity of handwritten signatures is required in various current life domains, notably with official contracts, banking or financial transactions. Therefore, in this paper a novel histogram-based descriptor and an improved classification of the bio-inspired Artificial Immune Recognition System (AIRS) are proposed for handwritten signature verification. Precisely, the Histogram Of Templates (HOT) is introduced to characterize the most widespread orientations of local strokes in handwritten signatures, while the combination of AIRS and SVM is proposed to achieve the verification task. Usually, using the k Nearest Neighbor rule, a questioned signature is classified by computing dissimilarities with respect to all AIRS outputs. In this work, using these dissimilarities, a second round of training is achieved by the SVM classifier to further improve the discrimination power. In comparison with existing methods, the experiments on two widely-used datasets show the potential and the effectiveness of the proposed system.

Soft-biometrics prediction from handwriting analysis is gaining a wide interest in writer identification since it gives additional knowledge about the writer like its gender (man or woman), its handedness (left-handed or right-handed) and its age range. All research works developed in this context were focused on predicting a single soft-biometric trait. Nevertheless, it could be more interesting to develop a system that predicts several traits from a handwritten text. Presently, we investigate the feasibility of such multiple trait prediction. To reach this end, we propose two prediction schemes. The first combines individual prediction scores to aggregate a global prediction. The second scheme is based on a multi-class prediction. For both schemes, the prediction is based on SVM classifier associated with Gradient features. Experimental corpus is collected from IAM handwritten database. Conclusively, the second scheme proved to be more promising and evinced that the age characteristic is stable over time for a certain category of writers.