Search

The direct current (DC) microgrid is one of the key research areas for our advancement toward carbon-free energy production. In this paper, a two-step controller is designed for the DC microgrid using a combination of the deep neural network (DNN) and exponential reaching law-based global terminal sliding mode control (ERL-GTSMC). The DC microgrid under consideration involves multiple renewable sources (wind, PV) and an energy storage unit (ESU) connected to a 700 V DC bus and a 4–12 kW residential load. The proposed control method eliminates the chattering phenomenon and offers quick reaching time by utilizing the exponential reaching law (ERL). In the two-step control configuration, first, DNNs are used to find maximum power point tracking (MPPT) reference values, and then ERL-based GTSMC is utilized to track the reference values. The real dynamics of energy sources and the DC bus are mathematically modeled, which increases the system's complexity. Through the use of Lyapunov stability criteria, the stability of the control system is examined. The effectiveness of the suggested hybrid control algorithm has been examined using MATLAB simulations. The proposed framework has been compared to traditional sliding mode control and terminal sliding mode control to showcase its superiority and robustness. Experimental tests based on the hardware-in-the-loop (HIL) setup are then conducted using 32-bit TMS320F28379D microcontrollers. Both MATLAB and HIL results show strong performance under a range of environmental circumstances and system uncertainties. [ABSTRACT FROM AUTHOR]

Human activity recognition (HAR) using wearable sensors enables continuous monitoring for healthcare applications. However, the conventional centralised training of deep learning models on sensor data poses challenges related to privacy, communication costs, and on-device efficiency. This paper proposes a federated learning framework integrating spiking neural networks (SNNs) with long short-term memory (LSTM) networks for energy-efficient and privacy-preserving HAR. The hybrid spiking-LSTM (S-LSTM) model synergistically combines the event-driven efficiency of SNNs and the sequence modelling capability of LSTMs. The model is trained using surrogate gradient learning and backpropagation through time, enabling fully supervised end-to-end learning. Extensive evaluations of two public datasets demonstrate that the proposed approach outperforms LSTM, CNN, and S-CNN models in accuracy and energy efficiency. For instance, the proposed S-LSTM achieved an accuracy of 97.36% and 89.69% for indoor and outdoor scenarios, respectively. Furthermore, the results also showed a significant improvement in energy efficiency of 32.30%, compared to simple LSTM. Additionally, we highlight the significance of personalisation in HAR, where fine-tuning with local data enhances model accuracy by up to 9% for individual users. [ABSTRACT FROM AUTHOR]

Digital holographic microscopy (DHM) is a valuable technique for investigating the optical properties of samples through the measurement of intensity and phase of diffracted beams. However, DHMs are constrained by Lagrange invariance, compromising the spatial bandwidth product (SBP) which relates resolution and field of view. Synthetic aperture DHM (SA-DHM) was introduced to overcome this limitation, but it faces significant challenges such as aberrations in synthesizing the optical information corresponding to the steering angle of incident wave. This paper proposes a novel approach utilizing deep neural networks (DNNs) for compensating aberrations in SA-DHM, extending the compensation scope beyond the numerical aperture (NA) of the objective lens. The method involves training a DNN from diffraction patterns and Zernike coefficients through a circular aperture, enabling effective aberration compensation in the illumination beam. This method makes it possible to estimate aberration coefficients from the only part of the diffracted beam cutoff by the circular aperture mask. With the proposed technique, the simulation results present improved resolution and quality of sample images. The integration of deep neural networks with SA-DHM holds promise for advancing microscopy capabilities and overcoming existing limitations. [ABSTRACT FROM AUTHOR]

Recent successes in deep learning have inspired researchers to apply deep neural networks to Acoustic Event Classification (AEC). While deep learning methods can train effective AEC models, they are susceptible to overfitting due to the models' high complexity. In this paper, we introduce EnViTSA, an innovative approach that tackles key challenges in AEC. EnViTSA combines an ensemble of Vision Transformers with SpecAugment, a novel data augmentation technique, to significantly enhance AEC performance. Raw acoustic signals are transformed into Log Mel-spectrograms using Short-Time Fourier Transform, resulting in a fixed-size spectrogram representation. To address data scarcity and overfitting issues, we employ SpecAugment to generate additional training samples through time masking and frequency masking. The core of EnViTSA resides in its ensemble of pre-trained Vision Transformers, harnessing the unique strengths of the Vision Transformer architecture. This ensemble approach not only reduces inductive biases but also effectively mitigates overfitting. In this study, we evaluate the EnViTSA method on three benchmark datasets: ESC-10, ESC-50, and UrbanSound8K. The experimental results underscore the efficacy of our approach, achieving impressive accuracy scores of 93.50%, 85.85%, and 83.20% on ESC-10, ESC-50, and UrbanSound8K, respectively. EnViTSA represents a substantial advancement in AEC, demonstrating the potential of Vision Transformers and SpecAugment in the acoustic domain. [ABSTRACT FROM AUTHOR]

Artificial intelligence (AI) radar technology offers several advantages over other technologies, including low cost, privacy assurance, high accuracy, and environmental resilience. One challenge faced by AI radar technology is the high cost of equipment and the lack of radar datasets for deep-learning model training. Moreover, conventional radar signal processing methods have the obstacles of poor resolution or complex computation. Therefore, this paper discusses an innovative approach in the integration of radar technology and machine learning for effective surveillance systems that can surpass the aforementioned limitations. This approach is detailed into three steps: signal acquisition, signal processing, and feature-based classification. A hardware prototype of the signal acquisition circuitry was designed for a Continuous Wave (CW) K-24 GHz frequency band radar sensor. The collected radar motion data was categorized into non-human motion, human walking, and human walking without arm swing. Three signal processing techniques, namely short-time Fourier transform (STFT), mel spectrogram, and mel frequency cepstral coefficients (MFCCs), were employed. The latter two are typically used for audio processing, but in this study, they were proposed to obtain micro-Doppler spectrograms for all motion data. The obtained micro-Doppler spectrograms were then fed to a simplified 2D convolutional neural networks (CNNs) architecture for feature extraction and classification. Additionally, artificial neural networks (ANNs) and 1D CNN models were implemented for comparative analysis on various aspects. The experimental results demonstrated that the 2D CNN model trained on the MFCC feature outperformed the other two methods. The accuracy rate of the object classification models trained on micro-Doppler features was 97.93%, indicating the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

A high cognitive load can overload a person, potentially resulting in catastrophic accidents. It is therefore important to ensure the level of cognitive load associated with safety-critical tasks (such as driving a vehicle) remains manageable for drivers, enabling them to respond appropriately to changes in the driving environment. Although electroencephalography (EEG) has attracted significant interest in cognitive load research, few studies have used EEG to investigate cognitive load in the context of driving. This paper presents a feasibility study on the simulation of various levels of cognitive load through designing and implementing four driving tasks. We employ machine learning-based classification techniques using EEG recordings to differentiate driving conditions. An EEG dataset containing these four driving tasks from a group of 20 participants was collected to investigate whether EEG can be used as an indicator of changes in cognitive load. The collected dataset was used to train four Deep Neural Networks and four Support Vector Machine classification models. The results showed that the best model achieved a classification accuracy of 90.37%, utilising statistical features from multiple frequency bands in 24 EEG channels. Furthermore, the Gamma and Beta bands achieved higher classification accuracy than the Alpha and Theta bands during the analysis. The outcomes of this study have the potential to enhance the Human–Machine Interface of vehicles, contributing to improved safety. [ABSTRACT FROM AUTHOR]

Keystroke dynamics is a soft biometric based on the assumption that humans always type in uniquely characteristic manners. Previous works mainly focused on analyzing the key press or release events. Unlike these methods, we explored a novel visual modality of keystroke dynamics for human identification using a single RGB-D sensor. In order to verify this idea, we created a dataset dubbed KD-MultiModal, which contains 243.2 K frames of RGB images and depth images, obtained by recording a video of hand typing with a single RGB-D sensor. The dataset comprises RGB-D image sequences of 20 subjects (10 males and 10 females) typing sentences, and each subject typed around 20 sentences. In the task, only the hand and keyboard region contributed to the person identification, so we also propose methods of extracting Regions of Interest (RoIs) for each type of data. Unlike the data of the key press or release, our dataset not only captures the velocity of pressing and releasing different keys and the typing style of specific keys or combinations of keys, but also contains rich information on the hand shape and posture. To verify the validity of our proposed data, we adopted deep neural networks to learn distinguishing features from different data representations, including RGB-KD-Net, D-KD-Net, and RGBD-KD-Net. Simultaneously, the sequence of point clouds also can be obtained from depth images given the intrinsic parameters of the RGB-D sensor, so we also studied the performance of human identification based on the point clouds. Extensive experimental results showed that our idea works and the performance of the proposed method based on RGB-D images is the best, which achieved 99.44 % accuracy based on the unseen real-world data. To inspire more researchers and facilitate relevant studies, the proposed dataset will be publicly accessible together with the publication of this paper. [ABSTRACT FROM AUTHOR]

The process of recognising and classifying radar signals and their radiation sources is currently a key element of operational activities in the electromagnetic environment. Systems of this type, called ELINT class systems, are passive solutions that detect, process, and analyse radio-electronic signals, providing distinctive information on the identified emission source in the final stage of data processing. The data processing in the mentioned types of systems is a very sophisticated issue and is based on advanced machine learning algorithms, artificial neural networks, fractal analysis, intra-pulse analysis, unintentional out-of-band emission analysis, and hybrids of these methods. Currently, there is no optimal method that would allow for the unambiguous identification of particular copies of the same type of radar emission source. This article constitutes an attempt to analyse radar signals generated by six radars of the same type under comparable measurement conditions for all six cases. The concept of the SEI module for the ELINT system was proposed in this paper. The main aim was to perform an advanced analysis, the purpose of which was to identify particular copies of those radars. Pioneering in this research is the application of the author's algorithm for the data particle geometrical divide, which at the moment has no reference in international publication reports. The research revealed that applying the data particle geometrical divide algorithms to the SEI process concerning six copies of the same radar type allows for almost three times better accuracy than a random labelling strategy within approximately one second. [ABSTRACT FROM AUTHOR]

The rapid growth in dataset sizes in modern deep learning has significantly increased data storage costs. Furthermore, the training and time costs for deep neural networks are generally proportional to the dataset size. Therefore, reducing the dataset size while maintaining model performance is an urgent research problem that needs to be addressed. Dataset condensation is a technique that aims to distill the original dataset into a much smaller synthetic dataset while maintaining downstream training performance on any agnostic neural network. Previous work has demonstrated that matching the training trajectory between the synthetic dataset and the original dataset is more effective than matching the instantaneous gradient, as it incorporates long-range information. Despite the effectiveness of trajectory matching, it suffers from complex gradient unrolling across iterations, which leads to significant memory and computation overhead. To address this issue, this paper proposes a novel approach called Expert Subspace Projection (ESP), which leverages long-range information while avoiding gradient unrolling. Instead of strictly enforcing the synthetic dataset's training trajectory to mimic that of the real dataset, ESP only constrains it to lie within the subspace spanned by the training trajectory of the real dataset. The memory-saving advantage offered by our method facilitates unbiased training on the complete set of synthetic images and seamless integration with other dataset condensation techniques. Through extensive experiments, we have demonstrated the effectiveness of our approach. Our method outperforms the trajectory matching method on CIFAR10 by 16.7% in the setting of 1 Image/Class, surpassing the previous state-of-the-art method by 3.2%. [ABSTRACT FROM AUTHOR]

Catastrophic forgetting, which means a rapid forgetting of learned representations while learning new data/samples, is one of the main problems of deep neural networks. In this paper, we propose a novel incremental learning framework that can address the forgetting problem by learning new incoming data in an online manner. We develop a new incremental learning framework that can learn extra data or new classes with less catastrophic forgetting. We adopt the hippocampal memory process to the deep neural networks by defining the effective maximum of neural activation and its boundary to represent a feature distribution. In addition, we incorporate incremental QR factorization into the deep neural networks to learn new data with both existing labels and new labels with less forgetting. The QR factorization can provide the accurate subspace prior, and incremental QR factorization can reasonably express the collaboration between new data with both existing classes and new class with less forgetting. In our framework, a set of appropriate features (i.e., nodes) provides improved representation for each class. We apply our method to the convolutional neural network (CNN) for learning Cifar-100 and Cifar-10 datasets. The experimental results show that the proposed method efficiently alleviates the stability and plasticity dilemma in the deep neural networks by providing the performance stability of a trained network while effectively learning unseen data and additional new classes. [ABSTRACT FROM AUTHOR]

This paper presents a generic framework for fault prognosis using autoencoder-based deep learning methods. The proposed approach relies upon a semi-supervised extrapolation of autoencoder reconstruction errors, which can deal with the unbalanced proportion between faulty and non-faulty data in an industrial context to improve systems' safety and reliability. In contrast to supervised methods, the approach requires less manual data labeling and can find previously unknown patterns in data. The technique focuses on detecting and isolating possible measurement divergences and tracking their growth to signalize a fault's occurrence while individually evaluating each monitored variable to provide fault detection and prognosis. Additionally, the paper also provides an appropriate set of metrics to measure the accuracy of the models, which is a common disadvantage of unsupervised methods due to the lack of predefined answers during training. Computational results using the Commercial Modular Aero Propulsion System Simulation (CMAPSS) monitoring data show the effectiveness of the proposed framework. [ABSTRACT FROM AUTHOR]

With the increasing demand for human-computer interaction and health monitoring, human behavior recognition with device-free patterns has attracted extensive attention. The fluctuations of the Wi-Fi signal caused by human actions in a Wi-Fi coverage area can be used to precisely identify the human skeleton and pose, which effectively overcomes the problems of the traditional solution. Although many promising results have been achieved, no survey summarizes the research progress. This paper aims to comprehensively investigate and analyze the latest applications of human behavior recognition based on channel state information (CSI) and the human skeleton. First, we review the human profile perception and skeleton recognition progress based on wireless perception technologies. Second, we summarize the general framework of precise pose recognition, including signal preprocessing methods, neural network models, and performance results. Then, we classify skeleton model generation methods into three categories and emphasize the crucial difference among these typical applications. Furthermore, we discuss two aspects, such as experimental scenarios and recognition targets. Finally, we conclude the paper by summarizing the issues in typical systems and the main research directions for the future. [ABSTRACT FROM AUTHOR]

Dealing with the islanded operation of a microgrid (MG), the micro sources must cooperate autonomously to regulate the voltage and frequency of the local power grid. Droop controller-based primary control is a method typically used to self-regulate voltage and frequency. The first problem of the droop method is that in a steady state, the microgrid's frequency and voltage deviate from their nominal values. The second concerns the power-sharing issue related to mismatched power line impedances between Distribution Generators (DGs) and MGs. A Secondary Control Unit (SCU) must be used as a high-level controller for droop-based primary control to address the first problem. This paper proposed a decentralized SCU scheme to deal with this issue using optimized PI controllers based on a Genetic Algorithm (GA) and Artificial Neural Networks (ANNs). The GA provides the appropriate adjustment parameters for all adopted PI controllers in the primary control-based voltage and current control loops and SCU-based voltage and frequency loops. ANNs are additionally activated in SCUs to provide precise online control parameter modification. In the proposed control structure, a virtual impedance method is adopted in the primary control scheme to address the power-sharing problem of parallel DGs. Further, in this paper, one of the main objectives includes electricity transmission over long distances using Low-Voltage DC Transmission (LVDCT) systems to reduce power losses and eradicate reactive power problems. Voltage Source Inverters (VSIs) are adopted to convert the DC electrical energy into AC near the consumer loads. The simulation results illustrated the feasibility of the proposed solutions in restoring voltage and frequency deviations, reducing line losses, as well as achieving active and reactive power sharing among the DGs connected to the MG. [ABSTRACT FROM AUTHOR]

Airborne radars are susceptible to a large number of clutter, noise and variable jamming signals in the real environment, especially when faced with active main lobe jamming, as the waveform shortcut technology in the traditional regime can no longer meet the actual battlefield radar anti-jamming requirements. Therefore, it is necessary to study anti-main-lobe jamming techniques for airborne radars in complex environments to improve their battlefield survivability. In this paper, we propose an airborne radar waveform design method based on a deep reinforcement learning (DRL) algorithm under clutter and jamming conditions, after previous research on reinforcement-learning (RL)-based airborne radar anti-jamming waveform design methods that have improved the anti-jamming performance of airborne radars. The method uses a Markov decision process (MDP) to describe the complex operating environment of airborne radars, calculates the value of the radar anti-jamming waveform strategy under various jamming states using deep neural networks and designs the optimal anti-jamming waveform strategy for airborne radars based on the duelling double deep Q network (D3QN) algorithm. In addition, the method uses an iterative transformation method (ITM) to generate the time domain signals of the optimal waveform strategy. Simulation results show that the airborne radar waveform designed based on the deep reinforcement learning algorithm proposed in this paper improves the signal-to-jamming plus noise ratio (SJNR) by 2.08 dB and 3.03 dB, and target detection probability by 26.79% and 44.25%, respectively, compared with the waveform designed based on the reinforcement learning algorithm and the conventional linear frequency modulation (LFM) signal at a radar transmit power of 5 W. The airborne radar waveform design method proposed in this paper helps airborne radars to enhance anti-jamming performance in complex environments while further improving target detection performance. [ABSTRACT FROM AUTHOR]

Medical images play an important role in medical diagnosis and treatment. Oncologists analyze images to determine the different characteristics of deadly diseases, plan the therapy, and observe the evolution of the disease. The objective of this paper is to propose a method for the detection of brain tumors. Brain tumors are identified from Magnetic Resonance (MR) images by performing suitable segmentation procedures. The latest technical literature concerning radiographic images of the brain shows that deep learning methods can be implemented to extract specific features of brain tumors, aiding clinical diagnosis. For this reason, most data scientists and AI researchers work on Machine Learning methods for designing automatic screening procedures. Indeed, an automated method would result in quicker segmentation findings, providing a robust output with respect to possible differences in data sources, mostly due to different procedures in data recording and storing, resulting in a more consistent identification of brain tumors. To improve the performance of the segmentation procedure, new architectures are proposed and tested in this paper. We propose deep neural networks for the detection of brain tumors, trained on the MRI scans of patients' brains. The proposed architectures are based on convolutional neural networks and inception modules for brain tumor segmentation. A comparison of these proposed architectures with the baseline reference ones shows very interesting results. MI-Unet showed a performance increase in comparison to baseline Unet architecture by 7.5 % in dice score, 23.91 % insensitivity, and 7.09 % in specificity. Depth-wise separable MI-Unet showed a performance increase by 10.83 % in dice score, 2.97 % in sensitivity, and 12.72 % in specificity as compared to the baseline Unet architecture. Hybrid Unet architecture achieved performance improvement of 9.71 % in dice score, 3.56 % in sensitivity, and 12.6 % in specificity. Whereas the depth-wise separable hybrid Unet architecture outperformed the baseline architecture by 15.45 % in dice score, 20.56 % in sensitivity, and 12.22 % in specificity. [ABSTRACT FROM AUTHOR]

Digitizing handwriting is mostly performed using either image-based methods, such as optical character recognition, or utilizing two or more devices, such as a special stylus and a smart pad. The high-cost nature of this approach necessitates a cheaper and standalone smart pen. Therefore, in this paper, a deep-learning-based compact smart digital pen that recognizes 36 alphanumeric characters was developed. Unlike common methods, which employ only inertial data, handwriting recognition is achieved from hand motion data captured using an inertial force sensor. The developed prototype smart pen comprises an ordinary ballpoint ink chamber, three force sensors, a six-channel inertial sensor, a microcomputer, and a plastic barrel structure. Handwritten data of the characters were recorded from six volunteers. After the data was properly trimmed and restructured, it was used to train four neural networks using deep-learning methods. These included Vision transformer (ViT), DNN (deep neural network), CNN (convolutional neural network), and LSTM (long short-term memory). The ViT network outperformed the others to achieve a validation accuracy of 99.05%. The trained model was further validated in real-time where it showed promising performance. These results will be used as a foundation to extend this investigation to include more characters and subjects. [ABSTRACT FROM AUTHOR]

Lane marking recognition is one of the most crucial features for automotive vehicles as it is one of the most fundamental requirements of all the autonomy features of Advanced Driver Assistance Systems (ADAS). Researchers have recently made promising improvements in the application of Lane Marking Detection (LMD). This research article has taken the initiative to review lane marking detection, mainly using deep learning techniques. This paper initially discusses the introduction of lane marking detection approaches using deep neural networks and conventional techniques. Lane marking detection frameworks can be categorized into single-stage and two-stage architectures. This paper elaborates on the network's architecture and the loss function for improving the performance based on the categories. The network's architecture is divided into object detection, classification, and segmentation, and each is discussed, including their contributions and limitations. There is also a brief indication of the simplification and optimization of the network for simplifying the architecture. Additionally, comparative performance results with a visualization of the final output of five existing techniques is elaborated. Finally, this review is concluded by pointing to particular challenges in lane marking detection, such as generalization problems and computational complexity. There is also a brief future direction for solving the issues, for instance, efficient neural network, Meta, and unsupervised learning. [ABSTRACT FROM AUTHOR]

This article presents a method for transparent watermarking of high-capacity watermarked video under H.265/HEVC (High-Efficiency Video Coding) compression conditions while maintaining high-quality encoded image. The aim of this paper is to present a method for watermark embedding using neural networks under conditions of subjecting video to lossy compression of the HEVC codec using the YUV420p color model chrominance channel for watermarking. This paper presents a method for training a deep neural network to embed a watermark when a compression channel is present. The discussed method is characterized by high accuracy of the video with an embedded watermark compared to the original. The PSNR (peak signal-to-noise ratio) values obtained are over 44 dB. The watermark capacity is 96 bits for an image with a resolution of 128 × 128. The method enables the complete recovery of a watermark from a single video frame compressed by the HEVC codec within the range of compression values defined by the CRF (constant rate factor) up to 22. [ABSTRACT FROM AUTHOR]

With many conveniences afforded by advances in smartphone technology, developing advanced data analysis methods for health-related information from smartphone users has become a fast-growing research topic in the healthcare field. Along these lines, this paper addresses smartphone sensor-based characterization of human motions with neural stochastic differential equations (NSDEs) and a Transformer model. NSDEs and modeling via Transformer networks are two of the most prominent deep learning-based modeling approaches, with significant performance yields in many applications. For the problem of modeling dynamical features, stochastic differential equations and deep neural networks are frequently used paradigms in science and engineering, respectively. Combining these two paradigms in one unified framework has drawn significant interest in the deep learning community, and NSDEs are among the leading technologies for combining these efforts. The use of attention has also become a widely adopted strategy in many deep learning applications, and a Transformer is a deep learning model that uses the mechanism of self-attention. This concept of a self-attention based Transformer was originally introduced for tasks of natural language processing (NLP), and due to its excellent performance and versatility, the scope of its applications is rapidly expanding. By utilizing the techniques of neural stochastic differential equations and a Transformer model along with data obtained from smartphone sensors, we present a deep learning method capable of efficiently characterizing human motions. For characterizing human motions, we encode the high-dimensional sequential data from smartphone sensors into latent variables in a low-dimensional latent space. The concept of the latent variable is particularly useful because it can not only carry condensed information concerning motion data, but also learn their low-dimensional representations. More precisely, we use neural stochastic differential equations for modeling transitions of human motion in a latent space, and rely on a Generative Pre-trained Transformer 2 (GPT2)-based Transformer model for approximating the intractable posterior of conditional latent variables. Our experiments show that the proposed method can yield promising results for the problem of characterizing human motion patterns and some related tasks including user identification. [ABSTRACT FROM AUTHOR]

The Internet of Things applications have become popular because of their lightweight nature and usefulness, which require low latency and response time. Hence, Internet of Things applications are deployed with the fog management layer (software) in closely located edge servers (hardware) as per the requirements. Due to their lightweight properties, Internet of Things applications do not consume many computing resources. Therefore, it is common that a small-scale data center can accommodate thousands of Internet of Things applications. However, in small-scale fog computing environments, task scheduling of applications is limited to offering low latency and response times. In this paper, we propose a latency-aware task scheduling method for Internet of Things applications based on artificial intelligence in small-scale fog computing environments. The core concept of the proposed task scheduling is to use artificial neural networks with partitioning capabilities. With the partitioning technique for artificial neural networks, multiple edge servers are able to learn and calculate hyperparameters in parallel, which reduces scheduling times and service level objectives. Performance evaluation with state-of-the-art studies shows the effectiveness and efficiency of the proposed task scheduling in small-scale fog computing environments while introducing negligible energy consumption. [ABSTRACT FROM AUTHOR]

In the recent past, hyper-spectral imaging has found widespread application in forensic science, performing both geometric characterization of biological traces and trace classification by exploiting their spectral emission. Methods proposed in the literature for blood stain analysis have been shown to be effectively limited to collaborative surfaces. This proves to be restrictive in real-case scenarios. The problem of the substrate material and color is then still an open issue for blood stain analysis. This paper presents a novel method for blood spectra correction when contaminated by the influence of the substrate, exploiting a neural network-based approach. Blood stains hyper-spectral images deposited on 12 different substrates for 12 days at regular intervals were acquired via a hyper-spectral camera. The data collected were used to train and test the developed neural network model. Starting from the spectra of a blood stain deposited in a generic substrate, the algorithm at first recognizes whether it is blood or not, then allows to obtain the spectra that the same blood stain, at the same time, would have on a reference white substrate with a mean absolute percentage error of 1.11%. Uncertainty analysis has also been performed by comparing the ground truth reflectance spectra with the predicted ones by the neural model. [ABSTRACT FROM AUTHOR]

Currently, Android is the most popular operating system among mobile devices. However, as the number of devices with the Android operating system increases, so does the danger of using them. This is especially important as smartphones increasingly authenticate critical activities(e-banking, e-identity). BotSense Mobile is a tool already integrated with some critical applications (e-banking, e-identity) to increase user safety. In this paper, we focus on the novel functionality of BotSense Mobile: the detection of malware applications on a user device. In addition to the standard blacklist approach, we propose a machine learning-based model for unknown malicious application detection. The lightweight neural network model is deployed on an edge device to avoid sending sensitive user data outside the device. For the same reason, manifest-related features can be used by the detector only. We present a comprehensive empirical analysis of malware detection conducted on recent data (May–June, 2022) from the Koodous platform, which is a collaborative platform where over 70 million Android applications were collected. The research highlighted the problem of machine learning model aging. We evaluated the lightweight model on recent Koodous data and obtained f 1 = 0.77 and high precision ( 0.9 ). [ABSTRACT FROM AUTHOR]

Imaging examinations are of remarkable importance for diagnostic support in Dentistry. Imaging techniques allow analysis of dental and maxillofacial tissues (e.g., bone, dentine, and enamel) that are inaccessible through clinical examination, which aids in the diagnosis of diseases as well as treatment planning. The analysis of imaging exams is not trivial; so, it is usually performed by oral and maxillofacial radiologists. The increasing demand for imaging examinations motivates the development of an automatic classification system for diagnostic support, as proposed in this paper, in which we aim to classify teeth as healthy or with endodontic lesion. The classification system was developed based on a Siamese Network combined with the use of convolutional neural networks with transfer learning for VGG-16 and DenseNet-121 networks. For this purpose, a database with 1000 sagittal and coronal sections of cone-beam CT scans was used. The results in terms of accuracy, recall, precision, specificity, and F1-score show that the proposed system has a satisfactory classification performance. The innovative automatic classification system led to an accuracy of about 70%. The work is pioneer since, to the authors knowledge, no other previous work has used a Siamese Network for the purpose of classifying teeth as healthy or with endodontic lesion, based on cone-beam computed tomography images. [ABSTRACT FROM AUTHOR]

Recent years evidenced an increase in the total number of skin cancer cases, and it is projected to grow exponentially. This paper proposes a computer-aided diagnosis system for the classification of a malignant lesion, where the acquired image is primarily pre-processed using novel methods. Digital artifacts such as hair follicles and blood vessels are removed, and thereafter, the image is enhanced using a novel method of histogram equalization. Henceforth, the pre-processed image undergoes the segmentation phase, where the suspected lesion is segmented using the Neutrosophic technique. The segmentation method employs a thresholding-based method along with a pentagonal neutrosophic structure to form a segmentation mask of the suspected skin lesion. The paper proposes a deep neural network base on Inception and residual blocks with softmax block after each residual block which makes the layer wider and easier to learn the key features more quickly. The proposed classifier was trained, tested, and validated over PH2, ISIC 2017, ISIC 2018, and ISIC 2019 datasets. The proposed segmentation model yields an accuracy mark of 99.50%, 99.33%, 98.56% and 98.04% for these datasets, respectively. These datasets are augmented to form a total of 103,554 images for training, which make the classifier produce enhanced classification results. Our experimental results confirm that the proposed classifier yields an accuracy score of 99.50%, 99.33%, 98.56%, and 98.04% for PH2, ISIC 2017, 2018, and 2019, respectively, which is better than most of the pre-existing classifiers. [ABSTRACT FROM AUTHOR]

Blind image deblurring is a challenging problem in computer vision, aiming to restore the sharp image from blurred observation. Due to the incompatibility between the complex unknown degradation and the simple synthetic model, directly training a deep convolutional neural network (CNN) usually cannot sufficiently handle real-world blurry images. An existed generative adversarial network (GAN) can generate more detailed and realistic images, but the game between generator and discriminator is unbalancing, which leads to the training parameters not being able to converge to the ideal Nash equilibrium points. In this paper, we propose a GAN with a dual-branch discriminator using multiple sparse priors for image deblurring (DBSGAN) to overcome this limitation. By adding the multiple sparse priors into the other branch of the discriminator, the task of the discriminator is more complex. It can balance the game between the generator and the discriminator. Extensive experimental results on both synthetic and real-world blurry image datasets demonstrate the superior performance of our method over the state of the art in terms of quantitative metrics and visual quality. Especially for the GOPRO dataset, the averaged PSNR improves 1.7 % over others. [ABSTRACT FROM AUTHOR]

An event camera is a novel bio-inspired sensor that effectively compensates for the shortcomings of current frame cameras, which include high latency, low dynamic range, motion blur, etc. Rather than capturing images at a fixed frame rate, an event camera produces an asynchronous signal by measuring the brightness change of each pixel. Consequently, an appropriate algorithm framework that can handle the unique data types of event-based vision is required. In this paper, we propose a dynamic object tracking framework using an event camera to achieve long-term stable tracking of event objects. One of the key novel features of our approach is to adopt an adaptive strategy that adjusts the spatiotemporal domain of event data. To achieve this, we reconstruct event images from high-speed asynchronous streaming data via online learning. Additionally, we apply the Siamese network to extract features from event data. In contrast to earlier models that only extract hand-crafted features, our method provides powerful feature description and a more flexible reconstruction strategy for event data. We assess our algorithm in three challenging scenarios: 6-DoF (six degrees of freedom), translation, and rotation. Unlike fixed cameras in traditional object tracking tasks, all three tracking scenarios involve the simultaneous violent rotation and shaking of both the camera and objects. Results from extensive experiments suggest that our proposed approach achieves superior accuracy and robustness compared to other state-of-the-art methods. Without reducing time efficiency, our novel method exhibits a 30 % increase in accuracy over other recent models. Furthermore, results indicate that event cameras are capable of robust object tracking, which is a task that conventional cameras cannot adequately perform, especially for super-fast motion tracking and challenging lighting situations. [ABSTRACT FROM AUTHOR]

A Microgrid (MG), like any other smart and interoperable power system, requires device-to-device (D2D) communication structures in order to function effectively. This communication system, however, is not immune to intentional or unintentional failures. This paper discusses the effects of communication link failures on MG control and management and proposes solutions based on enhancing message content to mitigate their detritus impact. In order to achieve this goal, generation and consumption forecasting using deep learning (DL) methods at the next time steps is used. The architecture of an energy management system (EMS) and an energy storage system (ESS) that are able to operate in coordination is introduced and evaluated by simulation tests, which show promising results and illustrate the efficacy of the proposed methods. It is important to mention that, in this paper, three dissimilar topics namely MG control/management, DL-based forecasting, and D2D communication architectures are employed and this combination is proven to be capable of achieving the aforesaid objective. [ABSTRACT FROM AUTHOR]

With the widespread adoption of service-oriented architectures (SOA), services with the same functionality but the different Quality of Service (QoS) are proliferating, which is challenging the ability of users to build high-quality services. It is often costly for users to evaluate the QoS of all feasible services; therefore, it is necessary to investigate QoS prediction algorithms to help users find services that meet their needs. In this paper, we propose a QoS prediction algorithm called the MFDK model, which is able to fill in historical sparse QoS values by a non-negative matrix decomposition algorithm and predict future QoS values by a deep neural network. In addition, this model uses a Kalman filter algorithm to correct the model prediction values with real-time QoS observations to reduce its prediction error. Through extensive simulation experiments on the WS-DREAM dataset, we analytically validate that the MFDK model has better prediction accuracy compared to the baseline model, and it can maintain good prediction results under different tensor densities and observation densities. We further demonstrate the rationality of our proposed model and its prediction performance through model ablation experiments and parameter tuning experiments. [ABSTRACT FROM AUTHOR]

In recent years, deep convolutional neural network (CNN)-based image enhancement has shown outstanding performance. However, due to the problems of uneven illumination and low contrast existing in endoscopic images, the implementation of medical endoscopic image enhancement using CNN is still an exploratory and challenging task. An endoscopic image enhancement network (EIEN) based on the Retinex theory is proposed in this paper to solve these problems. The structure consists of three parts: decomposition network, illumination correction network, and reflection component enhancement algorithm. First, the decomposition network model of pre-trained Retinex-Net is retrained on the endoscopic image dataset, and then the images are decomposed into illumination and reflection components by this decomposition network. Second, the illumination components are corrected by the proposed self-attention guided multi-scale pyramid structure. The pyramid structure is used to capture the multi-scale information of the image. The self-attention mechanism is based on the imaging nature of the endoscopic image, and the inverse image of the illumination component is fused with the features of the green and blue channels of the image to be enhanced to generate a weight map that reassigns weights to the spatial dimension of the feature map, to avoid the loss of details in the process of multi-scale feature fusion and image reconstruction by the network. The reflection component enhancement is achieved by sub-channel stretching and weighted fusion, which is used to enhance the vascular information and image contrast. Finally, the enhanced illumination and reflection components are multiplied to obtain the reconstructed image. We compare the results of the proposed method with six other methods on a test set. The experimental results show that EIEN enhances the brightness and contrast of endoscopic images and highlights vascular and tissue information. At the same time, the method in this paper obtained the best results in terms of visual perception and objective evaluation. [ABSTRACT FROM AUTHOR]

Random drift error is one of the important factors of MEMS (micro-electro-mechanical-system) sensor output error. Identifying and compensating sensor output error is an important means to improve sensor accuracy. In order to reduce the impact of white noise on neural network modeling, the ensemble empirical mode decomposition (EEMD) method was used to separate white noise from the original signal. The drift signal after noise removal is modeled by GRNN (general regression neural network). In order to achieve a better modeling effect, cross-validation and parameter optimization algorithms were designed to obtain the optimal GRNN model. The algorithm is used to model and compensate errors for the generated random drift signal. The results show that the mean value of original signal decreases from 0.1130 m/s2 to −1.2646 × 10−7 m/s2, while the variance decreases from 0.0133 m/s2 to 1.0975 × 10−5 m/s2. In addition, the displacement test was carried out by MEMS acceleration sensor. Experimental results show that the displacement measurement accuracy is improved from 95.64% to 98.00% by compensating the output error of MEMS sensor. By comparing the GA-BP (genetic algorithm-back propagation) neural network and the polynomial fitting method, the EEMD-GRNN method proposed in this paper can effectively identify and compensate for complex nonlinear drift signals. [ABSTRACT FROM AUTHOR]

The indoor navigation method shows great application prospects that is based on a wearable foot-mounted inertial measurement unit and a zero-velocity update principle. Traditional navigation methods mainly support two-dimensional stable motion modes such as walking; special tasks such as rescue and disaster relief, medical search and rescue, in addition to normal walking, are usually accompanied by running, going upstairs, going downstairs and other motion modes, which will greatly affect the dynamic performance of the traditional zero-velocity update algorithm. Based on a wearable multi-node inertial sensor network, this paper presents a method of multi-motion modes recognition for indoor pedestrians based on gait segmentation and a long short-term memory artificial neural network, which improves the accuracy of multi-motion modes recognition. In view of the short effective interval of zero-velocity updates in motion modes with fast speeds such as running, different zero-velocity update detection algorithms and integrated navigation methods based on change of waist/foot headings are designed. The experimental results show that the overall recognition rate of the proposed method is 96.77%, and the navigation error is 1.26% of the total distance of the proposed method, which has good application prospects. [ABSTRACT FROM AUTHOR]

Unlimited access to information and data sharing wherever and at any time for anyone and anything is a fundamental component of fifth-generation (5G) wireless communication and beyond. Therefore, it has become inevitable to exploit the super-high frequency (SHF) and millimeter-wave (mmWave) frequency bands for future wireless networks due to their attractive ability to provide extremely high data rates because of the availability of vast amounts of bandwidth. However, due to the characteristics and sensitivity of wireless signals to the propagation effects in these frequency bands, more accurate path loss prediction models are vital for the planning, evaluating, and optimizing future wireless communication networks. This paper presents and evaluates the performance of several well-known machine learning methods, including multiple linear regression (MLR), polynomial regression (PR), support vector regression (SVR), as well as the methods using decision trees (DT), random forests (RF), K-nearest neighbors (KNN), artificial neural networks (ANN), and artificial recurrent neural networks (RNN). RNNs are mainly based on long short-term memory (LSTM). The models are compared based on measurement data to provide the best fitting machine-learning-based path loss prediction models. The main results obtained from this study show that the best root-mean-square error (RMSE) performance is given by the ANN and RNN-LSTM methods, while the worst is for the MLR method. All the RMSE values for the given learning techniques are in the range of 0.0216 to 2.9008 dB. Furthermore, this work shows that the models (except for the MLR model) perform excellently in fitting actual measurement data for wireless communications in enclosed indoor environments since they provide R-squared and correlation values higher than 0.91 and 0.96 , respectively. The paper shows that these learning methods could be used as accurate and stable models for predicting path loss in the mmWave frequency regime. [ABSTRACT FROM AUTHOR]

Power distribution grids are typically installed outdoors and are exposed to environmental conditions. When contamination accumulates in the structures of the network, there may be shutdowns caused by electrical arcs. To improve the reliability of the network, visual inspections of the electrical power system can be carried out; these inspections can be automated using computer vision techniques based on deep neural networks. Based on this need, this paper proposes the Semi-ProtoPNet deep learning model to classify defective structures in the power distribution networks. The Semi-ProtoPNet deep neural network does not perform convex optimization of its last dense layer to maintain the impact of the negative reasoning process on image classification. The negative reasoning process rejects the incorrect classes of an input image; for this reason, it is possible to carry out an analysis with a low number of images that have different backgrounds, which is one of the challenges of this type of analysis. Semi-ProtoPNet achieves an accuracy of 97.22%, being superior to VGG-13, VGG-16, VGG-19, ResNet-34, ResNet-50, ResNet-152, DenseNet-121, DenseNet-161, DenseNet-201, and also models of the same class such as ProtoPNet, NP-ProtoPNet, Gen-ProtoPNet, and Ps-ProtoPNet. [ABSTRACT FROM AUTHOR]

This paper presents datasets utilised for synthetic near-infrared (NIR) image generation and bounding-box level fruit detection systems. A high-quality dataset is one of the essential building blocks that can lead to success in model generalisation and the deployment of data-driven deep neural networks. In particular, synthetic data generation tasks often require more training samples than other supervised approaches. Therefore, in this paper, we share the NIR+RGB datasets that are re-processed from two public datasets (i.e., nirscene and SEN12MS), expanded our previous study, deepFruits, and our novel NIR+RGB sweet pepper (capsicum) dataset. We oversampled from the original nirscene dataset at 10, 100, 200, and 400 ratios that yielded a total of 127 k pairs of images. From the SEN12MS satellite multispectral dataset, we selected Summer (45 k) and All seasons (180k) subsets and applied a simple yet important conversion: digital number (DN) to pixel value conversion followed by image standardisation. Our sweet pepper dataset consists of 1615 pairs of NIR+RGB images that were collected from commercial farms. We quantitatively and qualitatively demonstrate that these NIR+RGB datasets are sufficient to be used for synthetic NIR image generation. We achieved Frechet inception distances (FIDs) of 11.36, 26.53, and 40.15 for nirscene1, SEN12MS, and sweet pepper datasets, respectively. In addition, we release manual annotations of 11 fruit bounding boxes that can be exported in various formats using cloud service. Four newly added fruits (blueberry, cherry, kiwi and wheat) compound 11 novel bounding box datasets on top of our previous work presented in the deepFruits project (apple, avocado, capsicum, mango, orange, rockmelon and strawberry). The total number of bounding box instances of the dataset is 162 k and it is ready to use from a cloud service. For the evaluation of the dataset, Yolov5 single stage detector is exploited and reported impressive mean-average-precision, mAP [ 0.5 : 0.95 ] results of min:0.49, max:0.812. We hope these datasets are useful and serve as a baseline for future studies. [ABSTRACT FROM AUTHOR]

Recycling aluminium is essential for a circular economy, reducing the energy required and greenhouse gas emissions compared to extraction from virgin ore. A 'Twitch' waste stream is a mix of shredded wrought and cast aluminium. Wrought must be separated before recycling to prevent contamination from the impurities present in the cast. In this paper, we demonstrate magnetic induction spectroscopy (MIS) to classify wrought from cast aluminium. MIS measures the scattering of an oscillating magnetic field to characterise a material. The conductivity difference between cast and wrought makes it a promising choice for MIS. We first show how wrought can be classified on a laboratory system with 89.66% recovery and 94.96% purity. We then implement the first industrial MIS material recovery solution for sorting Twitch, combining our sensors with a commercial-scale separator system. The industrial system did not reflect the laboratory results. The analysis found three areas of reduced performance: (1) metal pieces correctly classified by one sensor were misclassified by adjacent sensors that only captured part of the metal; (2) the metal surface facing the sensor can produce different classification results; and (3) the choice of machine learning algorithm is significant with artificial neural networks producing the best results on unseen data. [ABSTRACT FROM AUTHOR]

Multiconnectivity allows user equipment/devices to connect to multiple radio access technologies simultaneously, including 5G, 4G (LTE), and WiFi. It is a necessity in meeting the increasing demand for mobile network services for the 5G and beyond wireless networks, while ensuring that mobile operators can still reap the benefits of their present investments. Multipath TCP (MPTCP) has been introduced to allow uninterrupted reliable data transmission over multiconnectivity links. However, energy consumption is a significant issue for multihomed wireless devices since most of them are battery-powered. This paper employs software-defined networking (SDN) and deep neural networks (DNNs) to manage the energy consumption of devices with multiconnectivity running MPTCP. The proposed method involves two lightweight algorithms implemented on an SDN controller, using a real hardware testbed of dual-homed wireless nodes connected to WiFi and cellular networks. The first algorithm determines whether a node should connect to a specific network or both networks. The second algorithm improves the selection made by the first by using a DNN trained on different scenarios, such as various network sizes and MPTCP congestion control algorithms. The results of our extensive experimentation show that this approach effectively reduces energy consumption while providing better network throughput performance compared to using single-path TCP or MPTCP Cubic or BALIA for all nodes. [ABSTRACT FROM AUTHOR]

This paper investigates the security–reliability of simultaneous wireless information and power transfer (SWIPT)-assisted amplify-and-forward (AF) full-duplex (FD) relay networks. In practice, an AF-FD relay harvests energy from the source (S) using the power-splitting (PS) protocol. We propose an analysis of the related reliability and security by deriving closed-form formulas for outage probability (OP) and intercept probability (IP). The next contribution of this research is an asymptotic analysis of OP and IP, which was generated to obtain more insight into important system parameters. We validate the analytical formulas and analyze the impact on the key system parameters using Monte Carlo simulations. Finally, we propose a deep learning network (DNN) with minimal computation complexity and great accuracy for OP and IP predictions. The effects of the system's primary parameters on OP and IP are examined and described, along with the numerical data. [ABSTRACT FROM AUTHOR]

Many "Industry 4.0" applications rely on data-driven methodologies such as Machine Learning and Deep Learning to enable automatic tasks and implement smart factories. Among these applications, the automatic quality control of manufacturing materials is of utmost importance to achieve precision and standardization in production. In this regard, most of the related literature focused on combining Deep Learning with Nondestructive Testing techniques, such as Infrared Thermography, requiring dedicated settings to detect and classify defects in composite materials. Instead, the research described in this paper aims at understanding whether deep neural networks and transfer learning can be applied to plain images to classify surface defects in carbon look components made with Carbon Fiber Reinforced Polymers used in the automotive sector. To this end, we collected a database of images from a real case study, with 400 images to test binary classification (defect vs. no defect) and 1500 for the multiclass classification (components with no defect vs. recoverable vs. non-recoverable). We developed and tested ten deep neural networks as classifiers, comparing ten different pre-trained CNNs as feature extractors. Specifically, we evaluated VGG16, VGG19, ResNet50 version 2, ResNet101 version 2, ResNet152 version 2, Inception version 3, MobileNet version 2, NASNetMobile, DenseNet121, and Xception, all pre-trainined with ImageNet, combined with fully connected layers to act as classifiers. The best classifier, i.e., the network based on DenseNet121, achieved a 97% accuracy in classifying components with no defects, recoverable components, and non-recoverable components, demonstrating the viability of the proposed methodology to classify surface defects from images taken with a smartphone in varying conditions, without the need for dedicated settings. The collected images and the source code of the experiments are available in two public, open-access repositories, making the presented research fully reproducible. [ABSTRACT FROM AUTHOR]

Temperature sensors are widely used in industrial production and scientific research, and accurate temperature measurement is crucial for ensuring the quality and safety of production processes. To improve the accuracy and stability of temperature sensors, this paper proposed using an artificial neural network (ANN) model for calibration and explored the feasibility and effectiveness of using ANNs to calibrate temperature sensors. The experiment collected multiple sets of temperature data from standard temperature sensors in different environments and compared the calibration results of the ANN model, linear regression, and polynomial regression. The experimental results show that calibration using the ANN improved the accuracy of the temperature sensors. Compared with traditional linear regression and polynomial regression, the ANN model produced more accurate calibration. However, overfitting may occur due to a small sample size or a large amount of noise. Therefore, the key to improving calibration using the ANN model is to design reasonable training samples and adjust the model parameters. The results of this study are important for practical applications and provide reliable technical support for industrial production and scientific research. [ABSTRACT FROM AUTHOR]

A spiking neural network (SNN) is a type of artificial neural network that operates based on discrete spikes to process timing information, similar to the manner in which the human brain processes real-world problems. In this paper, we propose a new spiking neural network (SNN) based on conventional, biologically plausible paradigms, such as the leaky integrate-and-fire model, spike timing-dependent plasticity, and the adaptive spiking threshold, by suggesting new biological models; that is, dynamic inhibition weight change, a synaptic wiring method, and Bayesian inference. The proposed network is designed for image recognition tasks, which are frequently used to evaluate the performance of conventional deep neural networks. To manifest the bio-realistic neural architecture, the learning is unsupervised, and the inhibition weight is dynamically changed; this, in turn, affects the synaptic wiring method based on Hebbian learning and the neuronal population. In the inference phase, Bayesian inference successfully classifies the input digits by counting the spikes from the responding neurons. The experimental results demonstrate that the proposed biological model ensures a performance improvement compared with other biologically plausible SNN models. [ABSTRACT FROM AUTHOR]

The use of higher frequency bands compared to other wireless communication protocols enhances the capability of accurately determining locations from ultra-wideband (UWB) signals. It can also be used to estimate the number of people in a room based on the waveform of the channel impulse response (CIR) from UWB transceivers. In this paper, we apply deep neural networks to UWB CIR signals for the purpose of estimating the number of people in a room. We especially focus on empirically investigating the various network architectures for classification from single UWB CIR data, as well as from various ensemble configurations. We present our processes for acquiring and preprocessing CIR data, our designs of the different network architectures and ensembles that were applied, and the comparative experimental evaluations. We demonstrate that deep neural networks can accurately classify the number of people within a Line of Sight (LoS), thereby achieving an 99% performance and efficiency with respect to both memory size and FLOPs (Floating Point Operations Per Second). [ABSTRACT FROM AUTHOR]

This paper presents a novel approach to reducing undesirable coupling in antenna arrays using custom-designed resonators and inverse surrogate modeling. To illustrate the concept, two standard patch antenna cells with 0.07λ edge-to-edge distance were designed and fabricated to operate at 2.45 GHz. A stepped-impedance resonator was applied between the antennas to suppress their mutual coupling. For the first time, the optimum values of the resonator geometry parameters were obtained using the proposed inverse artificial neural network (ANN) model, constructed from the sampled EM-simulation data of the system, and trained using the particle swarm optimization (PSO) algorithm. The inverse ANN surrogate directly yields the optimum resonator dimensions based on the target values of its S-parameters being the input parameters of the model. The involvement of surrogate modeling also contributes to the acceleration of the design process, as the array does not need to undergo direct EM-driven optimization. The obtained results indicate a remarkable cancellation of the surface currents between two antennas at their operating frequency, which translates into isolation as high as −46.2 dB at 2.45 GHz, corresponding to over 37 dB improvement as compared to the conventional setup. [ABSTRACT FROM AUTHOR]

Manufacturing systems need to be resilient and self-organizing to adapt to unexpected disruptions, such as product changes or rapid order, in supply chain changes while increasing the automation level of robotized logistics processes to cope with the lack of human experts. Deep Reinforcement Learning is a potential solution to solve more complex problems by introducing artificial neural networks in Reinforcement Learning. In this paper, a game engine was used for Deep Reinforcement Learning training, which allows visualization of view learning and result processes more intuitively than other tools, as well as a physical engine for a more realistic problem-solving environment. The present research demonstrates that a Deep Reinforcement Learning model can effectively address the real-time sequential 3D bin packing problem by utilizing a game engine to visualize the environment. The results indicate that this approach holds promise for tackling complex logistical challenges in dynamic settings. [ABSTRACT FROM AUTHOR]

This paper proposes a physics-informed neural network (PINN) for predicting the early-age time-dependent behaviors of prestressed concrete beams. The PINN utilizes deep neural networks to learn the time-dependent coupling among the effective prestress force and the several factors that affect the time-dependent behavior of the beam, such as concrete creep and shrinkage, tendon relaxation, and changes in concrete elastic modulus. Unlike traditional numerical algorithms such as the finite difference method, the PINN directly solves the integro-differential equation without the need for discretization, offering an efficient and accurate solution. Considering the trade-off between solution accuracy and the computing cost, optimal hyperparameter combinations are determined for the PINN. The proposed PINN is verified through the comparison to the numerical results from the finite difference method for two representative cross sections of PSC beams. [ABSTRACT FROM AUTHOR]

The study of Domain Generalization (DG) has gained considerable momentum in the Machine Learning (ML) field. Human Activity Recognition (HAR) inherently encompasses diverse domains (e.g., users, devices, or datasets), rendering it an ideal testbed for exploring Domain Generalization. Building upon recent work, this paper investigates the application of regularization methods to bridge the generalization gap between traditional models based on handcrafted features and deep neural networks. We apply various regularizers, including sparse training, Mixup, Distributionally Robust Optimization (DRO), and Sharpness-Aware Minimization (SAM), to deep learning models and assess their performance in Out-of-Distribution (OOD) settings across multiple domains using homogenized public datasets. Our results show that Mixup and SAM are the best-performing regularizers. However, they are unable to match the performance of models based on handcrafted features. This suggests that while regularization techniques can improve OOD robustness to some extent, handcrafted features remain superior for domain generalization in HAR tasks. [ABSTRACT FROM AUTHOR]

Pipeline structures are susceptible to corrosion, leading to significant safety, environmental, and economic implications. Existing long range guided wave inspection systems often fail to detect footprints of the concentrated defects, which can lead to leakage. One way to tackle this issue is the utilization of circumferential guided waves that inspect the pipe's cross section. However, achieving the necessary detection resolution typically necessitates the use of high-order modes hindering the inspection data interpretation. This study presents the implementation of an ultrasonic technique capable of detecting and classifying wall thinning and concentrated defects using high-order guided wave modes. The technique is based on a proposed phase velocity mapping approach, which generates a set of isolated wave modes within a specified phase velocity range. By referencing phase velocity maps obtained from defect-free stages of the pipe, it becomes possible to observe changes resulting from the presence of defects and assign those changes to the specific type of damage using artificial neural networks (ANN). The paper outlines the fundamental principles of the proposed phase velocity mapping technique and the ANN models employed for classification tasks that use synthetic data as an input. The presented results are meticulously verified using samples with artificial defects and appropriate numerical models. Through numerical modeling, experimental verification, and analysis using ANN, the proposed method demonstrates promising outcomes in defect detection and classification, providing a more comprehensive assessment of wall thinning and concentrated defects. The model achieved an average prediction accuracy of 92% for localized defects, 99% for defect-free cases, and 98% for uniform defects. [ABSTRACT FROM AUTHOR]

This paper presents a compact analog system-on-chip (SoC) implementation of a spiking neural network (SNN) for low-power Internet of Things (IoT) applications. The low-power implementation of an SNN SoC requires the optimization of not only the SNN model but also the architecture and circuit designs. In this work, the SNN has been constituted from the analog neuron and synaptic circuits, which are designed to optimize both the chip area and power consumption. The proposed synapse circuit is based on a current multiplier charge injector (CMCI) circuit, which can significantly reduce power consumption and chip area compared with the previous work while allowing for design scalability for higher resolutions. The proposed neuron circuit employs an asynchronous structure, which makes it highly sensitive to input synaptic currents and enables it to achieve higher energy efficiency. To compare the performance of the proposed SoC in its area and power consumption, we implemented a digital SoC for the same SNN model in FPGA. The proposed SNN chip, when trained using the MNIST dataset, achieves a classification accuracy of 96.56%. The presented SNN chip has been implemented using a 65 nm CMOS process for fabrication. The entire chip occupies 0.96 mm2 and consumes an average power of 530 μW, which is 200 times lower than its digital counterpart. [ABSTRACT FROM AUTHOR]

This paper presents a novel probabilistic machine learning (PML) framework to estimate the Brillouin frequency shift (BFS) from both Brillouin gain and phase spectra of a vector Brillouin optical time-domain analysis (VBOTDA). The PML framework is used to predict the Brillouin frequency shift (BFS) along the fiber and to assess its predictive uncertainty. We compare the predictions obtained from the proposed PML model with a conventional curve fitting method and evaluate the BFS uncertainty and data processing time for both methods. The proposed method is demonstrated using two BOTDA systems: (i) a BOTDA system with a 10 km sensing fiber and (ii) a vector BOTDA with a 25 km sensing fiber. The PML framework provides a pathway to enhance the VBOTDA system performance. [ABSTRACT FROM AUTHOR]

Human activity recognition (HAR) is essential for the development of robots to assist humans in daily activities. HAR is required to be accurate, fast and suitable for low-cost wearable devices to ensure portable and safe assistance. Current computational methods can achieve accurate recognition results but tend to be computationally expensive, making them unsuitable for the development of wearable robots in terms of speed and processing power. This paper proposes a light-weight architecture for recognition of activities using five inertial measurement units and four goniometers attached to the lower limb. First, a systematic extraction of time-domain features from wearable sensor data is performed. Second, a small high-speed artificial neural network and line search method for cost function optimization are used for activity recognition. The proposed method is systematically validated using a large dataset composed of wearable sensor data from seven activities (sitting, standing, walking, stair ascent/descent, ramp ascent/descent) associated with eight healthy subjects. The accuracy and speed results are compared against methods commonly used for activity recognition including deep neural networks, convolutional neural networks, long short-term memory and convolutional–long short-term memory hybrid networks. The experiments demonstrate that the light-weight architecture can achieve a high recognition accuracy of 98.60%, 93.10% and 84.77% for seen data from seen subjects, unseen data from seen subjects and unseen data from unseen subjects, respectively, and an inference time of 85 μ s. The results show that the proposed approach can perform accurate and fast activity recognition with a reduced computational complexity suitable for the development of portable assistive devices. [ABSTRACT FROM AUTHOR]

In fringe projection profilometry, high-order harmonics information of distorted fringe will lead to errors in the phase estimation. In order to solve this problem, a point-wise phase estimation method based on a neural network (PWPE-NN) is proposed in this paper. The complex nonlinear mapping relationship between the gray values and the phase under non-sinusoidal distortion is constructed by using the simple neural network model. It establishes a novel implicit expression for phase solution without complicated measurement operations. Compared with the previous method of combining local image information, it can accurately calculate each phase value by point. The comparison results show that the traditional method is with periodic phase errors, while the proposed method can effectively eliminate phase errors caused by non-sinusoidal phase shifting. [ABSTRACT FROM AUTHOR]