Traditionally, speech emotion recognition (SER) research has relied on manually handcrafted acoustic features using feature engineering. However, the design of handcrafted features for complex SER tasks requires significant manual effort, which impedes generalisability and slows the pace of innovation. This has motivated the adoption of representation learning techniques that can automatically learn an intermediate representation of the input signal without any manual feature engineering. Representation learning has led to improved SER performance and enabled rapid innovation. Its effectiveness has further increased with advances in deep learning (DL), which has facilitated deep representation learning where hierarchical representations are automatically learned in a data-driven manner. This paper presents the first comprehensive survey on the important topic of deep representation learning for SER. We highlight various techniques, related challenges and identify important future areas of research. Our survey bridges the gap in the literature since existing surveys either focus on SER with hand-engineered features or representation learning in the general setting without focusing on SER.

This paper presents a methodology that uses a software development approach for designing processors in a higher programming language; Chisel and demonstrates a completely open-source route for fabricating the processor.

I have made some revisions according to the advices of other researchers and submitted a new version of the theory (v4). I also have added a supplementary material for the the new version, including some detailed explanations and derivations. In the proposed theory, the total electromagnetic energy of a radiator is separated into three parts: a Coulomb-velocity energy, a radiative energy, and a macroscopic Schott energy. The Coulomb-velocity energy is considered to be attached to the sources as the same in the charged particle theory. It becomes zero as soon as its sources have disappeared. The radiative energy leaves the radiator and propagates to the surrounding space. The macroscopic Schott energy continues to exist for a short time after the sources have disappeared. It is a kind of oscillating energy and is considered to be responsible for energy exchange between the reactive energy and the radiative energy, performing like the Schott energy in the charged particle theory. As the Poynting vector describes the total power flux density related to the total electromagnetic energy, it should include the contributions of the real radiative power and a pseudo power flow caused by the fluctuation of the reactive energy. The energies involved in the electromagnetic mutual coupling are interpreted in a similar way. In the theory, all energies are defined with explicit expressions in which the vector potential plays an important role. The time domain formulation and the frequency domain formulation of the theory are in consistent with each other. The theory is also verified with Hertzian dipole. Numerical examples demonstrate that the theory may provide insightful interpretation for electromagnetic radiation and mutual coupling problems.

Artificial neural network (ANN) ability to learn, correct errors, and transform a large amount of raw data into useful medical decisions for treatment and care have increased its popularity for enhanced patient safety and quality of care. Therefore, this paper reviews the critical role of ANNs in providing valuable insights for patients’ healthcare decisions and efficient disease diagnosis. We thoroughly review different types of ANNs presented in the existing literature that advanced ANNs adaptation for complex applications. Moreover, we also investigate ANN’s advances for various disease diagnoses and treatments such as viral, skin, cancer, and COVID-19. Furthermore, we propose a novel deep Convolutional Neural Network (CNN) model called ConXNet for improving the detection accuracy of COVID-19 disease. ConXNet is trained and tested using different datasets, and it achieves more than 97% detection accuracy and precision, which is significantly better than existing models. Finally, we highlight future research directions and challenges such as complexity of the algorithms, insufficient available data, privacy and security, and integration of biosensing with ANNs. These research directions require considerable attention for improving the scope of ANNs for medical diagnostic and treatment applications.

Although the skills required to solve isolated robotics problems are reaching amazing performances recently, we propose the evaluation of such individual solutions in fully integrated robot systems tested in real daily situations like those presented at international robotics competitions. The simulation Domestic Standard Platform League (sDSPL), which utilizes the HSR simulator developed for the World Robot Summit, surges from the necessity to standardise and spread the research on Domestic Service Robots where a series of solutions can be tested to solve a general-purpose task in a standard domestic environment; this approach has been proven successful at several international competitions, namely, the RoboCup Japan Open, the Mexican Tournament of Robotics, and the RoboCup 2021. **This work was accepted to The 39th Annual Conference of the Robotics Society of Japan.

A document by Yuwei Sun . Click on the document to view its contents.

Electro Encephalo Gram (EEG) is a monitoring method used in biomedical and computer science to understand brain activity. Therefore, the analysis and classification of these signals play a prominent role in estimating a person’s behavior to certain events. Manually analyzing these signals is very tedious and time-consuming, so an automated scientific tool is required to analyze the brain signals. In this work, the authors are explored various pre-processing segmentation techniques that are helpful in an automatic machine and deep learning-based classification methods available for EEG signal processing. Most of the machine and deep learning methods are followed pre-processing as a common step in classification. Extraction of the basic sub-band components from EEG signals such as delta (δ), theta (θ), alpha (α), beta (β), and gamma (γ) is very important in the pre-processing stage. These sub bands of EEG signal have extraordinary evidence related to multiple neurophysiological processes, which are useful for further prediction & diagnosis of diseases and other emotion-based applications. This review paper elaborates various elementary ideas of extracting EEG sub-bands and the role of EEG in Brain-Computer Interface (BCI) in the classification. (Submitted To IEEE reviews in Biomedical Engineering)

A novel two-electrode, frequency-scan electrical impedance tomography (EIT) system for gesture recognition not only reduces the measurement complexity and the number of electrodes, but also achieves a high accuracy in recognizing common gestures and pinch gestures. A bespoke circuit with two medical electrodes was developed to collect data from the back of the hand and presented a frequency-scan method to increase the diversity of impedance data. The data were processed using data cleaning and feature extraction methods. The processed data were then sent to machine learning classification models for training and realizing accurate gesture recognition. To verify the effectiveness of this system, we designed two groups of nine gestures in a hand-gesture recognition experiment. The results showed that the system can achieve a recognition accuracy of 98.3% with a group of four common gestures and an accuracy of 97.4% with a group of five pinch gestures. Additionally, two proof-of-concept interactive scenarios were implemented to demonstrate the general purpose of this system.

The quality of the technical textiles, in particular nonwovens, depends on the process parameters involved in producing them. It is not practical to investigate the effect of these parameters in real-time. Hence, a digital twin has been developed to analyze nonwoven production processes. The quality of the virtually produced nonwovens from the digital twin is then analyzed and mapped back to the process parameters. Different visualization techniques can be incorporated to guide engineers in finding optimal combinations of these process parameters for obtaining the desired product quality. This would eliminate the traditional trial and error strategies and reduce the domain expertise required in such an analysis. In this paper, we present a visual analytic tool that aids the engineer in analyzing and optimizing the quality of the nonwovens, thereby helping to make beneficial decisions in the industrial production process.

Horses and breeders need to be safe on the farm and the riding club. On account of the great value of the horse, the breeder needs to protect it from theft and disease. In this context, it is important to detect and to recognize the identity of each horse for security reasons. In fact, this paper proposes a Smart Riding Club Biometric System (SRCBS) consisting in automatically detecting and recognizing horses as well as humans. The proposed system is based on the facial biometrics for a horse and the gait biometrics for a human due to their simplicity and intuitiveness in an uncontrolled environment. The present work focuses mainly on horse face detection and recognition. Animal face detection is still extremely difficult given the fact that face textures and shapes are grossly diverse. In addition, recent detectors require a huge dataset for training and represent a huge number of parameters and layers, leading to so much training time. For resolving these problems and also for a useful detection system, this paper proposes a Sparse Neural Network (SNN) based on sparse features for horse face detection. Different global and local features were performed to identify horses and humans for the recognition process. Due to the unavailability of horse databases, this paper presents a new benchmark for horse detection and recognition in order to evaluate our proposed system. This system achieved an average precision equal to 90% for horse face detection and a recognition rate equal to 99.89% for horse face identification.

In this study, product of two independent and non-identically distributed (i.n.i.d.) random variables (RVs) for κ-μ fading distribution and α-μ fading distribution is considered. The novel exact series formulas for the product of two i.n.i.d. fading distributions κ-μ and α-μ are derived instead of Fox H-function to solve the problem that Fox H function with multiple RVs cannot be implemented in professional mathematical software packages as MATHEMATICA and MAPLE. Exact close-form expressions of probability density function (PDF) and cumulative distribution function (CDF) are deduced to represent provided product expressions and generalized composite multipath shadowing models. At last, these analytical results are validated with Monte Carlo simulations, it shows that for provided κ-μ/α-μ model nonlinear parameter has more important influence than multipath component in PDF and CDF when the ratio between the total power of the dominant components and the total power of the scattered waves is same.

Electrified urban railway systems are large consumers of energy in urban areas and thus, there is a need for energy saving measures in this transportation sector. Recuperation of train’s regenerative braking energy (RBE) is one of the best ways for attaining high levels of energy efficiency in this area. Energy Storage Systems (ESSs) prove to be the most practical and viable solution for maximizing the RBE utilization in urban railway systems. In the existing related reviews of ESSs, few papers covered the review of ESS technologies or converter interface. However, a comprehensive review is needed in this regard. This paper discusses an overview of urban railway electrification, and detail review for the three ESS components – ESS Technologies, Bidirectional DC-DC Converters (BDCs), and Controller Unit. This study concludes that among the storage technologies, supercapacitor ESS appears to be the most suitable followed by Lithium-ion batteries and flywheels. For BDC, cascaded BDC is the most suitable followed by dual active bridge. For control methods, fuzzy logic and artificial intelligent are recommended among other control strategies. Thus, the key contribution of this paper is the comprehensive review and analyses of the ESS’s components in the recovery of RBE in urban railways.

This paper reviews sensorless algorithms for both induction motors and permanent magnet motors using the active flux model, such that any design applicable for non-salient pole ac motors can also be included in the review framework. The proposed review framework classifies all sensorless algorithms following a five-layer hierarchy abbreviated as O-I-M-A-I, resulting in four main categories as i) inherently sensorless position estimation, ii) non-inherently sensorless position estimation, iii) post-position-estimation speed estimation, and iv) speed estimation for indirect field orientation. Various ac motor models are derived by assuming a constant active flux amplitude, based on which seven generic sensorless algorithms are summarized in a tutorial. Recommendations are made for sensorless drive designers to begin with inherently sensorless method such that the two-way coupling between position estimation and speed estimation is avoided. Finally, classical induction motor model results from time-varying active flux amplitude and slip relation, for which a state transformation is recommended for achieving global stability.

Anomaly detection aims to identify the true anomalies from a given set of data instances. Unsupervised anomaly detection algorithms are applied to an unlabeled dataset by producing a ranked list based on anomaly scores. Unfortunately, due to the inherent limitations, many of the top-ranked instances by unsupervised algorithms are not anomalies or not interesting from an application perspective, which leads to high false-positive rates. Active anomaly discovery (AAD) is proposed to overcome this deficiency, which sequentially selects instances to get the labeling information and incorporate it into the anomaly detection algorithm to improve the detection accuracy iteratively. However, labeling is often costly. Therefore, the way to balance detection accuracy and labeling cost is essential. Along this line, this paper proposes a novel AAD method to achieve the goal. Our approach is based on the state-of-the-art unsupervised anomaly detection algorithm, namely, Isolation Forest, to extract features. Thereafter, the sparsity of the extracted features is utilized to improve its anomaly detection performance. To enforce the sparsity of the features and subsequent improvement of the detection analysis, a new algorithm based on online gradient descent, namely, Sparse Approximated Linear Anomaly Discovery (SALAD), is proposed with its theoretical Regret analysis. Extensive experiments on both open-source and additive manufacturing datasets demonstrate that the proposed algorithm significantly outperforms the state-of-the-art algorithms for anomaly detection.

If reality is being augmented as a simulation; then from the modern norm of physics, it is possible that everywhere around us, including us and what we perceive is a simulation simulated by a super powerful computer from a farfetched future times, taken the Einstein’an notion, that, past, present and future occur simultaneously. It is quite probable between two consecutive amplitudes that, the simulation can either be done by the super advanced civilization in greater than or equal to Kardashev 3.0 scale or by some higher order entities existing in a dimensional domain beyond our perceiveness and notion of our understanding. To agree to the fact of the simulation hypothesis, there exists a mathematical foundation of the desired logic behind this simulation, which will be investigated throughout this paper whose another consequence might be the déjà vu or the Mandela effect. The errors arise in this simulation is a form of glitch in the matrix that should happen because of the commutable lagging of the super-intelligent computers of either future ones or higher-order ones. Preciseness about the calculations of dimensions opened a way for t + s = 2 + 10 where the non-locality of the time being perceived as a 2-dimensional entity opens up the door for further investigations. The more will be discussed in detail in this paper.

The pervasiveness of online mis/disinformation escalated during the COVID-19 pandemic. To address the proliferation of online mis/disinformation, it is critical to build reliability into the tools older adults use to seek health information. On average, older adult populations demonstrate disproportionate susceptibility to false messages spread under the guise of accuracy and were the most engaged with false information about COVID-19 across online platforms when compared to other age-groups. In a design-thinking challenge posed by AARP to graduate students in a Digital Health course at Tufts University School of Medicine, students leveraged existing solutions to design a web browser extension that is responsive to both passive and active health information-seeking methods utilized by older adults in the United States. This paper details the design-thinking process employed, insights gained from primary research, an overview of the prototyped solution, and insights relating to the design of effective health information-seeking platforms for older adults.

Prosthetic control for rehabilitation, among many other applications, can leverage in-sensor hand gesture recognition in which lightweight machine learning models for classifying electromyogram (EMG) signals are embedded on miniature, low-power devices. While research efforts have demonstrated high accuracy in controlled settings, these methods have yet to make a significant commercial or clinical impact due to the wide variety of scenarios and situational contexts that are faced during everyday use. Typical static models suffer from the effects of EMG signal variation caused by changing contexts in which they are deployed. Here, we propose an incremental learning algorithm using hyperdimensional (HD) computing that can efficiently learn gesture patterns performed in new limb positions, a context-change which normally significantly degrades classification accuracy. A prototype-based learning algorithm, HD computing enables memory- and computation-efficient incorporation of new training examples into the model, while preserving information about already learned contexts. We present various configurations of the incremental HD classifier, allowing system designers to trade classification performance for implementation efficiency as measured through memory footprint. Incremental learning experiments with data from 5 subjects show that HD computing can achieve similar accuracies as incrementally trained SVM and LDA classifiers while requiring a fraction of the memory allocation.

Walking gait data measured using force platforms is a promising means for person re-identification in authentication and surveillance scenarios. We aimed to determine the most discriminant components of force platform data using a two-stream Convolutional Recurrent Neural Network (KineticNet). Each network in the two-stream architecture extracts features pertaining to a single stance phase and then these features are fused to represent the entire gait cycle. Over two sessions, ground reaction forces (Fx, Fy, Fz), moments (Mx, My, Mz), and center of pressure coordinates (Cx, Cy) were acquired from 118 participants as they walked our laboratory five times at preferred speed. For each participant and each session, up to three samples were reserved for network training, leaving one sample for network validation and one sample for network testing. KineticNet’s performance was evaluated using both individual component and multi-component inputs before ablation studies were conducted on its architecture. Fz was the most discriminant individual component, and re-identification using Fz, Fy, and Cy together was the most accurate overall at 96.02%. These results warrant further investigation into the utility of force platforms as an accessory or alternative to video cameras for gait based person re-identification.