Keywords: Technology Assessment
Abstract: Presentation of the five best paper candidates (selected by the IEEE SMC Best Conference Paper Awards Subcommittee):

• Candidate 1 (Mo-PS4-T11.1): "A Method for Evaluation and Optimization of Automotive Camera Systems Based on Simulated Raw Sensor Data" (Roos, Stefan)
• Candidate 2 (Tu-PS3-T3.4): "What Comes after Telepresence? Embodiment, Social Presence and Transporting One’s Functional and Social Self" (van Erp, Jan)
• Candidate 3 (We-PS3-T5.3): "Learning Temporal Context of Normality for Unsupervised Anomaly Detection in Videos" (Hyun, Wooyeol)
• Candidate 4 (Mo-PS1-T2.3): "SeeWay: Vision-Language Assistive Navigation for the Visually Impaired" (Yang, Zongming)
• Candidate 5 (We-PS1-T10.5): "Human Intracortical Responses to Varying Electrical Stimulation Conditions Are Separable in Low-Dimensional Subspaces" (Sun, Samantha)

Keywords: Multimedia Systems, Information Visualization, Human-Computer Interaction
Abstract: Human action recognition (HAR) is a popular topic in developing a visual analysis system because of its tremendous potential in autonomous visual analysis. However, visual analysis is a sophisticated field in computer vision because an image sequence consists of various features that do not belong to a specific action. Therefore, we present a novel architecture approach for human action recognition and localization. We dubbed it DTA, an abbreviation of the detect, track, and analyze. It is inspired by yolov3, deep-sort, and 3D convolutional neural networks. Our framework is compact in analyzing human action, and the results showed that the proposed method outperforms previous state-of-the-art methods in various aspects. Moreover, the action recognition model is developed, trained, and tested using the ROI version of the KTH dataset. The experimental results showed the accuracy of the proposed model is superior compared to other traditional methods.

Keywords: Multimedia Systems, Information Visualization, Information Systems for Design
Abstract: Reading news is an important way for people to obtain information. People can quickly sort out the context of events through a short news video. However, there are numerous news generated around the world every day. It's challenge to locate the interesting video. Thumbnails are often used as video covers and play an important role in displaying video content and driving views. Video owners can choose from individual images or elaborate thumbnails to upload to the site. But manually selecting from a large number of frames is time-consuming, and customizing thumbnails requires a high degree of expertise. Therefore, this paper proposes an automatic generation method of news video thumbnail, which can screen out semantically similar contents according to user query and combine them into a thumbnail. In order to facilitate the screening of graphic materials, we also propose a video content structuring method based on multiple cues, which can accurately segment the video into theme units. At the same time, we designed a visual system to display thumbnails, and designed a user survey to investigate the performance of this method in news retrieval and understanding. Compared with peer methods, the thumbnails generated by our method can help users better understand the video content and locate the videos they are interested in.

Keywords: Team Performance and Training Systems, Entertainment Engineering, Information Visualization
Abstract: The term “esports” is used to refer to game-based competitions considered as sports events. Esports currently involves a qualitative instruction that relies on individual experience and intuition, such as instruction from professional gamers. By contrast, sports observe a shift from qualitative to high-quality instruction using quantitative evaluation indices. Therefore, in esports, evaluation indices must be quantified to improve the training quality. This study focuses on crossing situations in soccer games. Crossing is a play in which a player passes from the side to the goal. It is important in soccer because of its high possibility of leading to a goal. Therefore, the quality of cross training in soccer must be improved to increase the scoring probability by crossing. As a feasibility study, this work aims at a quantitative evaluation of crossing in soccer games. The important factors for a successful cross are the positional relationship and the speed of each player at the time of the cross. A crossing scene is modeled to obtain a cross score based on these factors. First, when the velocity and trajectory of the cross ball are determined, the area is defined in which the offensive and defensive players can touch the ball. This area is calculated, and the value of how close to the center the ball trajectory passes in relation to the area is determined for each player. Using these values, the Cross Score (CS) was obtained by constructing a cross evaluation formula. To confirm the validity of the obtained evaluation values, a system that can obtain the evaluation values using image processing was developed. Then, 264 videos of crossing scenes in a soccer game were obtained, and CS was calculated for each of them. The correlation coefficient between the cross score and the percentage of successful crosses is 0.923, showing a strong correlation and confirming CS validity. As an example of application of the CS, it is possible to judge whether a cross is good or bad by displaying the CS on a heatmap using quantitative values. This enables visual and quantitative feedback using CS, suggesting the possibility of improving the quality of training.

Keywords: Human-Computer Interaction, Multimedia Systems
Abstract: The decoupled head for classification and localization have been proven powerful in the most of one-stage and two-stage detectors. However, most object detection algorithm share a Feature Pyramid Networks. We perform a thorough analysis about the effectiveness of Feature Pyramid Networks for these two tasks. The decoupled feature pyramid network performs better than the shared network. Going a step further, we found that the two tasks have different preferences for feature pyramid networks. For higher accuracy, we propose a Scene Parsing Pyramid Network for Classification and a Feature Pyramid Transformer Network for Localization. Scene Parsing Pyramid Network exploit the capability of global context information by different region based context aggregation through pyramid pooling module and pyramid attention feature extraction module. Feature Pyramid Transformer Network can capture the suitable contexts of objects residing in different scales. We evaluate our double Feature Pyramid Networks feature pyramid network in the object detection task by integrating it into the FCOS algorithm. The modified algorithm outperforms previous state-of-the-art feature pyramid based methods with a clear margin on both MS-COCO 2017 validation and test datasets.

Keywords: Human-Computer Interaction, Multimedia Systems
Abstract: Generative Adversarial Networks (GANs) have achieved great improvement after using Convolutional Neural Networks (CNNs) instead of Multi-Layer Perceptrons (MLPs) to build network architecture. Recently, since Transformer architecture has performed well in compute vision, building a Transformer-based image generation network helps solve some of problems caused by CNNs e.g. CNNs-based GANs are difficult to train. On the other hand, the learning of positional encoding in the Transformer structure is often ignored in Transformer-based GANs. Capsule networks are usually considered to be able to learn position information in image features. Therefore, this paper constructs a Progressive Growing Transformer network with Capsule Embedding GAN (PGTCEGAN). The results from the proposed approach are promising with 3.59 FID and 3.92 FID on CelebA and LSUN-Church datasets respectively in image generation task.

Keywords: Decision Support Systems, Intelligent Power Grid, Intelligent Green Production Systems
Abstract: For better decision-making on the combined heat and power economic emission dispatch (CHPEED), this paper proposes a novel algorithm by enhancing the multi-objective evolutionary algorithm based on decomposition (MOEA/D), a representative multi-objective evolutionary algorithm with promising performance but not wildly adopted in optimizing the complex economic emission dispatch. Firstly, utilizing the Tchebycheff aggregate objective function (AOF) weighted with a group of evenly distributed weight vectors, the proposed algorithm transforms the original CHPEED problem approximately into a series of single-objective sub-problems. Given this, the solutions, i.e., the Pareto optimal solutions(POS), corresponding to each weighted subproblem respectively, represent the corresponding tradeoff preference between the fuel cost and the emission. Such tradeoff preferences are helpful for decision-making in practical engineering. Besides, in the proposed algorithm, an effective constrained optimization strategy is hybridized to meet the equality constraint, which leads to promising results in the simulations to encourage further studies.

Keywords: Decision Support Systems
Abstract: The effective allocation of personnel avoids the waste of resources and enables the organization to achieve the maximum possible results. In this work, we present a decision aid for Navy personnel allocation based on the Coppe-Cosenza Model and the Hungarian Algorithm. We implemented the model in R, using existing packages, and can be operated by non-experts. The proposed model extends the potentialities of the Coppe-Cosenza Model, presenting an optimal cost solution and presenting a feasible transformation to use the aggregation matrix as a basis for a loss/cost function. Finally, we build a feasible and meaningful integration of both techniques to adapt them to the problem domain peculiarities. Since the optimization algorithm is polynomial, the model can be used for considerable large problems.

Keywords: Decision Support Systems, Distributed Intelligent Systems, Smart Buildings, Smart Cities and Infrastructures
Abstract: Spatio-temporal prediction has a wide range of applications in many fields, e.g., air pollution, weather forecasting, and traffic forecasting. Water quality prediction is also one of spatio-temporal prediction tasks. However, it faces the following challenges: 1) Water quality in river networks has complex spatial dependencies; 2) There are complex nonlinear relations in water quality time series; and 3) It is difficult to realize long-term forecasting. To address these challenges, this work proposes a spatio-temporal prediction model called a Graph Attention-based Spatio-Temporal (GAST) neural network. GAST investigates spatial and temporal dependencies of water quality time series. First, we introduce a temporal attention mechanism to capture time series dependencies, which can effectively handle nonlinear relationships in time series. Second, we adopt a spatial attention mechanism to extract spatial dependencies of river networks and fuse temporal features of spatial nodes. Third, we adopt a temporal convolution residual mechanism based on the spatio-temporal fusion, which improves the accuracy of long-term series prediction. This work adopts two real-world datasets to evaluate the proposed GAST and experiments demonstrate that GAST outperforms several state-of-the-art methods in terms of prediction accuracy.

Keywords: Decision Support Systems, Modeling of Autonomous Systems, Control of Uncertain Systems
Abstract: Abstract—To seize the development opportunity of artificial intelligence and deepen the supply-side structural reform of talent cultivation, this study focuses on the integration of the traditional advantage engineering specialty with artificial intelligence. Six key factors affecting the integration of traditional engineering and artificial intelligence were identified by the structural equation model. Based on the innovative problem- solving theory (TRIZ), the key factors were transformed into six key problems, and the TRIZ problem model and solution model were established by taking the key problems as standard factors. Then, with the help of TRIZ contradiction matrix and 40 invention and innovation principle tools, we provide solutions to problems and transform them into professional integration strategies. Finally, it can be an auxiliary decision support tool for the integration strategy-making of traditional engineering specialty and artificial intelligence. Through this tool, decision-makers can understand each link of the decision-making process more clearly, and make the strategy-making process more scientific and closer to the real needs.

Keywords: Decision Support Systems, Quality and Reliability Engineering
Abstract: Problem solving in complex situations or systems requires an appropriate and reliable understanding of the problem. We propose a framework that combines the Iceberg and Situational Awareness models. The goal is to raise decision makers' awareness so that, depending on the system layers, problems can be perceived as events and trends within the system structure. Depending on what has been perceived, decision makers are then able to understand the impact of problems or their underlying causes and conditions. The appropriateness of the perceived and integrated information depends on whether it helps decision makers answer their questions and whether it corresponds to reality. Reliability depends on the completeness and certainty of the information presented. The perception of system properties is influenced by the (in)transparency of the real system and the prevailing paradigm and mindset of the observer. So, we propose a system taxonomy within the framework. This summarizes system properties of simple, complicated and complex systems and the resulting behavioral characteristics identified in the literature. Decision makers can use the taxonomy to explore the properties of real systems to develop an appropriate system model and assess its reliability, or to evaluate existing system models whether, or to what extent, they are suitable for decision support. This is because, based on these models, decision makers develop a corresponding understanding of the problem, upon which their decisions for possible solutions are based.

Keywords: Quality and Reliability Engineering, Service Systems and Organizations
Abstract: The storage overhead of blocks and storage load of nodes are increasing as consortium blockchains grow. The existing storage models consider either low-reliability issues or high storage overhead but not both. This work proposes a new storage model that combines group-based block storage with block splitting and unit encoding for the first time. The former is used to improve the reliability of a block, while the latter is adopted to reduce the storage overhead and increase the reliability of a block in a consortium blockchain. This work analyzes the storage overhead of blocks, storage load of nodes, and reliability of a consortium blockchain and designs a storage optimization method to reduce the storage overhead and improve the reliability of a blockchain system. Experimental results show that the proposed method outperforms Multi-layer Practical Byzantine Fault Tolerance, RapidChain, and Byzantine Fault Tolerance-Store in terms of storage overhead and reliability.

Keywords: Robotic Systems, Distributed Intelligent Systems, Modeling of Autonomous Systems
Abstract: Unmanned aerial vehicles (UAVs) have been very successful in many civilian and commercial applications, including disaster relief, search and rescue, precision farming, archaeology, cargo transport, and surveillance. In most of these applications, mapping is a required phase that needs to be performed as an initial step. While mapping has attracted much attention in the last decades, much of the works rely on single drones. In this context, we propose a multiple UAV system for efficient mapping, minimizing mission time and cost. The system includes offline and online planning, and a good balance between both to reduce on-board processing. Offline planning includes area decomposition, take-off location finding, and path planning. Online planning will then be used to react to any unforeseen event that might occur during the plan execution. These incidents include a sudden change in weather conditions, communication loss or drone malfunction, and the presence of a nearby flying obstacle. Each of the main offline and online planning tasks are formalized as a Multi-Objective Optimization (MOO) problem where requirements need to be met while objectives have to be optimized. In this regard, we consider several evolutionary techniques to tackle these MOO problems. To assess the performance of these techniques, we conducted several experiments and reported the related results. One finding is that MOEA/D outperforms NSGA2, while the latter requires less processing time.

Keywords: Smart Metering, System Modeling and Control, Technology Assessment
Abstract: Inside medical environments, hand hygiene is very important to avoid healthcare-associated infections. The traditional method to measure hand sanitization guidelines consists of having an external observer watching the medical personnel, which can introduce bias and can only make the observations for a fixed amount of time. Some alternatives have been proposed, using different sensors and machine learning to replace the role of the observer, placing many sensors, or assuming that hand hygiene occurs in a fixed place, like a gel dispenser. In this paper, we report on our approach to detection and tracking of medical staff, as well as measuring their hand disinfection compliance using minimal equipment, obtaining only 2D images from one camera to avoid cluttering hospital space or change the personnel’s regular behavior, achieving promising results for this task in this setup.

Keywords: Consumer and Industrial Applications, Cooperative Systems and Control
Abstract: The general context of this paper is the development and the application of the guaranteed state estimation observer–based interval techniques, to improve the navigation unit used in space missions. A H∞ constraint is also considered in the design of the interval observer, to formulate robustness performance against sensor misalignment errors, noises and other unknown inputs that may affect the estimation. The application support is the Microscope satellite which is a scientific mission launched in 2016. A functional engineering simulator (FES) of the Microscope mission is used to assess the performance of the proposed state estimation interval techniques. The FES includes highly representative models of sensors and actuators, and Dynamics Kinematics and Environment (DKE) models. The environment modules (within DKE) contain the spatial disturbances that affect the rotational and translational dynamics of the satellite. The considered disturbances are the magnetic field, the aerodynamic drag, the gravitational disturbances, the solar and the albedo radiations.

The paper focuses on the guaranteed estimation of the attitude angles. The proposed interval observer–based estimation algorithm is implemented within the navigation unit of the Microscope mission. It is followed by a data fusion algorithm to deliver the estimates to the control unit. It is shown in the paper how the fusion rule parameters can be determined for the proposed fusion algorithm to be optimal in the l1-norm sense.

A comparison is provided in the paper with i) the hybridation filter originally implemented in the Microscope navigation unit, and ii), two other interval approaches recently reported in the literature. Advantages and limitations are discussed based on the results obtained from a simulation campaign conducted with the Microscope’s FES.

A last and fundamental discussion is finally conducted about the existence of a separation principle about the overall control scheme.

Results obtained from a simulation campaign conducted with the Microscope FES, demonstrate the potential of the proposed approach.

Keywords: Technology Assessment
Abstract: In recent decades, technological advances have exponentially accelerated the search for digital transformation. The use of technologies based on Artificial Intelligence (AI) is essential to achieve these goals. In this work, we investigate the future trends in the field. We conducted a systematic review and analyzed their content with a methodology based on Future-Oriented Technology Analysis (FTA) techniques. We have identified 30 future events related to digital transformation and AI and their respective feasible years. Our findings are categorized into industry fields: business process management, construction industry, digital economy, human resources, healthcare, industry, social innovation, tourism, and education. Our work allows decision-makers to draw up strategies, map the paths to be taken, and better prepare for their future needs, using AI as a tool for value aggregation.

Keywords: Swarm Intelligence, Evolutionary Computation
Abstract: Artificial intelligence (AI) has fast developed nowadays especially in the deep learning field that most of the deep neural networks pursue the better problem-solving ability by making the network models deeper, larger, and more complex. However, too large and too complex AI algorithms/models require too large computational burden, which is not reality in academic researches nor the right way to real human intelligence. Moreover, in another research branch of AI named evolutionary computation (EC), which is inspired by the biological evolution of nature and swarm intelligence behaviours, will the EC algorithms become more efficient with larger and more complex algorithm model when solving complicated optimization problems like the large-scale optimization problems (LSOPs)? To this concern, this paper investigates whether some existing large-scale optimization EC algorithms can further improve their performance in solving LSOPs by only increasing the population size. We select 12 representative algorithms for investigation, including 4 standard EC algorithms and 8 well-known large-scale optimization algorithms. Then, we adopt the widely-used IEEE Congress on Evolutionary Computation (CEC 2010) LSOPs benchmark test suite to compare the performance of the same algorithms with different population sizes. The experimental results show that simply increasing the population size does not necessarily improve the performance of algorithms in solving LSOPs.

Keywords: Neural Networks and their Applications, Evolutionary Computation, Deep Learning
Abstract: Recently, Graph Neural Networks (GNNs) have shown great promise in addressing various tasks with non-Euclidean data. Encouraged by the successful application on discovering convolutional and recurrent neural networks, Neural Architecture Search (NAS) is extended to alleviate the complexity of designing appropriate task-specific GNNs. Unfortunately, existing graph NAS methods are usually susceptible to unscalable depth, redundant computation, constrained search space and some other limitations. In this paper, we present an evolutionary graph neural network architecture search strategy, involving inheritance, crossover and mutation operators based on fine-grained atomic operations. Specifically, we design two novel crossover operators at different granularity levels, GNNCross and LayerCross. Experiments on three different graph learning tasks indicate that the neural architectures generated by our method exhibit comparable performance to the handcrafted and automated baseline GNN models.

Keywords: Evolutionary Computation, Swarm Intelligence, Heuristic Algorithms
Abstract: We propose an adaptive scouting strategy to further improve the performance of fireworks algorithm (FWA) by perfecting the scouting strategy, where firework individuals find and track potential searching directions by generating spark individuals one by one instead of generating all at once within their explosion amplitude. The new strategy makes two modifications to the previous strategy to better apply to various optimization scenarios. The first difference is that the initial explosion center migrates with better generated offspring individuals (i.e., spark individuals) instead of being fixed, i.e., the next round of the initial explosion center will return to the recently generated spark individuals instead of initial firework individuals when the current tracing direction has no potential. The other modification is to adaptively adjust the explosion amplitude of subsequent explosion operations according to current search results. When better spark individuals are generated, the explosion amplitude is reduced to enhance the local search ability. Otherwise, expand the explosion amplitude to escape from trapped local areas quickly. To evaluate the performance of our newly proposed strategy, we use exactly the same parameter settings with that used in the first proposed literature. The experimental results confirmed that the adaptive scouting strategy shows better performance and faster convergence speed especially for complex multimodal optimization problems.

Keywords: Evolutionary Computation, Heuristic Algorithms
Abstract: Gradient-based repair aims to repair infeasible solutions to feasible ones using the gradient information of the constraints. As an effective constraint handling method, gradient-based repair has received extensive attention and has been applied in various evolutionary algorithms (EAs). Nevertheless, due to the complexity of constraints in practical problems, a single infeasible solution often needs to be repaired multiple times until it becomes a feasible solution or reaches the maximum number of repairs. As far as we know, existing related research on gradient-based repair mainly applies this method directly to EAs, while there is little work in the evolutionary computing community on how to improve gradient-based repair. Currently, the multiple repairs for a single individual are independent. That is, the current repair does not consider the previous repair experience. However, only using gradient information to repair infeasible individuals may result in oscillations in the search process. Therefore, in this paper, we propose a heuristic gradient-based repair method (HGR) which exploits the previous repair information of an individual to alleviate this issue. Experimental results on several benchmarks demonstrate the effectiveness of the proposed method.

Keywords: Computational Intelligence, Swarm Intelligence, Evolutionary Computation
Abstract: Metaheuristics are crucial for solving complex optimization problems effectively in many fields. With the recent increased interest in explainable artificial intelligence (XAI), the issue of how metaheuristics interact with XAI has attracted much attention. Most metaheuristics have stochastic search processes, and the jaguar algorithm (JA) is a unique algorithm that has exact search paths. JA shows potential abilities both in exploration and exploitation, but still faces limitations. Therefore, this study proposes a new metaheuristic based on JA and invents precision hunting behavior (PH-JA), which inherits the concept of JA and significantly strengthens its search efficiency. PH-JA improves the solution quality by adaptively detecting the current environment's trends during movement and precisely hunting prey. Then, PH-JA makes the best of the historical information to reduce the computational cost. PH-JA is an explainable metaheuristic and has systematic operational procedures in that each movement is performed according to the present situation rather than random elements. Therefore, the path to a solution is interpretable and can return the same result with the same input. The experimental results demonstrated the superiority of PH-JA, which is better than other classical and state-of-the-art metaheuristics in terms of solution quality and evaluation costs.

Keywords: Cybernetics for Informatics, Cloud, IoT, and Robotics Integration, Optimization and Self-Organization Approaches
Abstract: In recent years, “security by design” has become a fundamental requirement for new technologies, especially for those that interconnect various systems and their components, e.g., implanted medical devices with remote configurability capabilities. As such, security of implanted medical devices has become a concern exacerbated by the tight constraints characterizing them and the need to ensure patients’ safety as these devices typically work to assure life-critical functions. To protect such systems, multiple proposals have been made to detect threats and mitigate vulnerabilities and malware intrusions. A common approach is utilizing hardware redundancy to deal with system faults, however, implanted medical devices work under tight power and area constraints that may not always allow for redundancy. In this paper, we build on an existing multi-modal based mitigation scheme for securing medical devices and propose self-healing capabilities that would allow the device to automatically reestablish functionalities compromised due to external threats.

Keywords: Application of Artificial Intelligence, Cybernetics for Informatics, Computational Life Science
Abstract: Electromyography-based control applications became popular in recent days since they can be implemented using cheap brain-computer interface devices. In the past, for classifying electromyogram data, human experts would extract features from raw signals manually. Unfortunately, this is a time-consuming process and requires extensive domain knowledge. With the advancement of artificial intelligence, such procedures can be done automatically using deep learning architectures. In this study, we designed a deep convolutional network to classify electromyogram signals. Furthermore, a support vector machine-based classifier was also implemented as a reference system. The goal of this study was to develop a control architecture, in which the user interface is equipped with a cheap commercial brain-computer interface system for mobile robot control. We collected a considerable amount of electromyogram signals during different facial expressions that can be used as intervention signals. Finally, we implemented an online real-time control method based on the deep convolutional network. The system enables the user to drive a mobile robot platform solely with electromyogram signals. Manipulating the designed system, after a short initial practice period, the users could navigate through a predefined path with a low error rate.

Keywords: Application of Artificial Intelligence, Biometric Systems and Bioinformatics, Computational Intelligence
Abstract: One of the most challenging area of diabetes research is to provide such automated insulin delivery systems -- so called artificial pancreas systems -- that have robust and adaptive capabilities in a highly sophysticated way. I.e. they are able to provide robust insulin delivery actions at the beginning of the therapy to satisfy the requirements of the patients without knowing the users daily lifestyle and preferences however adaptive on the short-term to learn these patient specifics to increase the quality of the therapy. One possible solution is the closed-loop systems that have self-learning features. In the present study, we have examined a glucose regulatory problem using direct reinforcement learning based controller. The approach represents the fully automatic insulin administration as the timepoint and the carbohydrate content of the meals were unknown and randomized. We constructed a virtual environment of the patient with type 1 diabetes by applying a mathematical model. Proximal policy optimization learning model with continuous action space was used. Furthermore, we evaluated the effect of different training lengths on the test scenario.

Keywords: Computational Intelligence, Cybernetics for Informatics, Heuristic Algorithms
Abstract: Linear regression is one of the fundamental tasks of mathematical statistics and machine learning related disciplines. Several techniques have been elaborated, however the most common one is still the application of the least squares technique. In this paper we present an alternative, ''robustified'' approach for linear regression that instead of relying on the minimization of the L2-norm, minimizes the P-norm that was introduced by Steiner et. al. in connection with earth sciences and textit{Most Frequent Value} (MFV) calculations. Although the proposed alternative serves with a rather robust and outlier-resistant regression, a nonlinear equation system has to be solved in an iterative way. The present study serves with a comparison between the Newton's and Broyden's method solving the nonlinear system in the iterative procedure in case of a test example and in case of real life economic data regarding unconditional economic β-convergence of the EU countries and regions.

Keywords: Deep Learning, Neural Networks and their Applications, Knowledge Acquisition
Abstract: Biomedical event extraction is a fundamental information extraction task, which aims to identify biomedical event triggers and parameters in the text. Although various deep learning and Graph Convolutional Network (GCN) models have been proposed for this task, these models are insufficient to acquire enough local and global context information of documents. To effectively extract joint local and global context information, we propose a joint biomedical event extraction model named BGHGCN, which consists of Bi-directional Long Short-Term Memory (BiLSTM), improved BiAffine Graph Parser (IBGP), GCN and hypergraph convolutional networks (HGCN). Our model employs BiLSTM to learn word sequence features and uses improved BiAffine Graph Parser to enrich dependent syntax features. Afterwards we use GCN to extract local features from IBGP and BiLSTM. Specifically, we introduce HGCN to jointly extract local and global context information with a new fusion mechanism of local feature and incidence matrix, which can effectively extract structural features of hypergraph including node and hyperedge features. Finally, we evaluated our model on two biomedical event datasets MLEE and GE to compare with other baseline models.

Keywords: Neural Networks and their Applications, Application of Artificial Intelligence, Deep Learning
Abstract: Human-assist robots are designed to reduce the burden on the body when lifting objects and assist the elderly, disabled, and other people with muscle weakness in their daily activities. They also avoid accidents and assist motions by recognizing what kind of motion the user is about to make. Therefore, human-assist systems need to estimate the motion intention of a user in real time. The sooner the assist robot can accurately recognize the user’s motion, the sooner the assist robot can plan which motion to assist and how to assist it to guarantee success. If the user’s intended motion and its phase are estimated in the early stage of the motion, the assist robot can figure out how the user is moving by comparing with standard motion models to assist the motion as necessary. This paper proposes a method to estimate the user's intended motion and its phase simultaneously in real time based on integrated information consisting of the user’s posture, motion, EMG signals, and the surrounding environment. Two kinds of artificial neural networks are applied in the proposed method. Damping neurons are used in the artificial neural network to estimate the motion phase effectively. The intended lower-limb motions and their phases in daily living motion are estimated in real-time. The effectiveness of the proposed method was evaluated by performing experiments of lower-limb motion.

Keywords: Deep Learning
Abstract: Baseball elbow is a pitching elbow disorder caused by repeated pitching movements. Osteochondritis dissecans (OCD) is one of baseball elbow disorders, and is an intractable osteochondral injury that tends to occur in elementary and junior high school students. If it can be found in the early stages, it will be completely cured by conservative treatment, which is to set a period to stop playing baseball. Since there is almost no pain in the early stage, the hurdles for consultation are high and there are many cases in which the condition becomes severe. Periodical medical check of baseball elbow is effective, however, the number of implementations is several times a year due to the shortage of specialists who can make a diagnosis. In this study, for the purpose of developing computer-aided diagnosis (CADx) of early-stage OCD, we propose an OCD detection method using ultrasound images of the elbow. The proposed method first segments the humerus capitellum using fully convolutional network (FCN). Secondly, the segmented region is classified into OCD +/- classes using fine-tuning VGG16 to detect OCD. The proposed method was applied to 125 child baseball players including 61 OCD children and 64 healthy children. 5-fold cross-validation was conducted. The average detection results were 76.8% for accuracy, 100% for precision, 52.3% for recall, F1-score was 0.673, and AUC was 0.851.

Keywords: Deep Learning, Image Processing and Pattern Recognition, Transfer Learning
Abstract: Chronic lung disease of the newborn (CLD) is the most common and serious lung disease in premature infants and tends to cause developmental problems in adulthood, including brain, vision, and hearing as well as cardiopulmonary function. Premature infants and very small premature infants admitted to the NICU are often born without sufficient lung function, and chest X-rays are frequently taken to confirm the state of lung function. Although prior studies have been conducted to predict CLD severity using patient information in premature infants, prediction using chest X-ray images has not been performed. In this study, we predict the severity (mild or severe) of neonatal chest X-ray images using Convolutional Neural Networks (CNNs) to enable early intervention to provide personalized treatment and improve prognosis. Thirty subjects were tested in a leave-one-out cross validation experiment using 30 chest X-ray images of 11 patients with mild disease and 19 patients with severe disease at 7 days of age. To improve the prediction accuracy, we proposed to limit the input image of the CNN to the lung field region and to use a pre-trained model for transfer learning. Automatic lung field area extraction using YOLOv5, one of the object recognition models with a real-time object detection algorithm, showed a reliability of more than 0.82 for all chest X-ray images, and pre-training using chest X-ray images predicted pneumonia with an accuracy of 0.880. Then, four different experiments were conducted, comparing the results with different input images (whole image or lung field region) and with and without transition learning. The results showed that the best accuracy was obtained when the entire image was used as input and no transition learning was performed, with an Accuracy of 0.667. Future work includes creating a learning model that takes into account unbalanced data, changing parameters, and improving prediction accuracy by combining the features obtained from the chest X-ray images of newborns in this study with individual patient information, such as gender and gestational age.

Keywords: Cybernetics for Informatics, Media Computing
Abstract: We proposed an interactive map application, OTASCE Map, which presents map and route information. It utilizes customizable audio-tactile elements when the user traces the map on the touchscreen or the displayed route with their finger. However, despite the unique interaction on the OTASCE Map, the quantitative interaction situation and the operation requirements under screen-reading conditions remain unclear. Therefore, in this study, we aimed to clarify the actual situation of route-tracing interaction on the smartphone-based map application and to derive efficiency improvement factors. At the exhibition for the visually impaired in Japan, we evaluated the traceability of the route on the OTASCE Map. The result indicated that most participants could significantly trace the route and reach the destination when they did not perform release-retouch-trace (RRT) interaction. Herein, we also discuss the cause of RRT interaction and how to decrease such interactions.

Keywords: Application of Artificial Intelligence, Image Processing and Pattern Recognition, Deep Learning
Abstract: Federated Learning (FL) enables multiple institutes to train models collaboratively without sharing private data. Most of the current FL research focuses on perspectives such as communication efficiency, privacy protection, and personalization. Almost all work assumed that the data of FL are already ideally collected. However, in medical image analysis scenarios, data annotation demands both expertise and tedious labor, which means it is a critical problem that cannot be neglected in FL. In this study, we proposed a federated active learning (FedAL) framework that can decrease the annotation workload while maintaining the performance of FL. To the best of our knowledge, this is the first federated active learning framework working on medical images. Using only up to 50% of samples, our FedAL was able to achieve state-of-the-art performance on the real-world dermoscopic task. Our FedAL outperformed active learning methods under FL and achieved the performance comparable to full data FL.

Keywords: System Modeling and Control, Mechatronics, Robotic Systems
Abstract: Quantization errors occur when a signal is measured by digital sensors that have naturally a limited accuracy. When applying position control to a large scale robot, these quantization errors cause steps in the control input that possibly excite structural oscillations. To prevent this phenomenon, a state observer is designed that provides smooth estimated states with high accuracy for systems with quantized output. It is realized by observing the error dynamics of the measurement signal. The resulting observer is updated at each time instance where the quantized value of the measured signal changes. By using a corrected value of the output as well as an orthogonal projection, the updated estimation error leads to smooth estimated states. The design of this triggered observer does not have more design parameter than a conventional Luenberger observer and is therefore readily to implement. However, the overall estimation error is smaller in a direct comparison to such a Luenberger observer which is shown in simulations using a model of a mechanical drive system.

Keywords: System Modeling and Control
Abstract: This paper introduces a fast learning mechanism to address constrained input in the higher derivatives Newton-based extremum seeking. The proposed algorithm has a two-time-scale structure consisting of: a compensation mechanism (i.e. an anti-windup compensator) with fast dynamics that compensates for the effect of the constrained input, and a slow subsystem to maximize the map’s higher derivatives by regulating the output. The practical asymptotic stability of the new ES algorithm is proved using a modified version of the singular perturbation method. The effectiveness of the proposed algorithm is demonstrated using simulations.

Keywords: System Modeling and Control
Abstract: In this paper, we study modeling of air traffic flow from flight trajectory data in the airspace. Compared to road/train traffic, modeling of air traffic flow is difficult because trajectory of each aircraft fluctuates in the 3-dimensional space due to various factors such as weather and congestion level. By this reason, finding important routes on which aircrafts frequently pass is necessary for building air traffic flow models. We propose a framework for finding important routes in the form of graphs based on combination of various technologies such as space partition, trajectory clustering, and skeleton extraction.

Keywords: System Modeling and Control, Mechatronics, Cooperative Systems and Control
Abstract: The concept of FAIR data (Findable, Accessible, Interoperable and Reusable) is well-known in orthopedic research. With this notion in mind, the validated UGent Knee Rig (UGKR) is extended to facilitate ground reaction force (GRF) control. The bench mark system in dynamic knee rigs is the Oxford Knee Rig which is the basis of many knee rigs in literature. It has the intrinsic quality of imposing a constant ground reaction force on a specimen. A new control strategy is introduced to extend the functionality of the UGKR with GRF control. It consist of a model-based control strategy with feedback and gain adaptation based on model identification. The presented control strategies are tested on mechanical hinges and cadaver specimens to verify its performance. The results show that the presented control strategies are capable of meeting the control specification of a force error less than 10 N as reported in literature. This innovation allows for comparison between groups working on dynamic knee rigs on an international level resulting in a more unified platform for data comparison.

Keywords: System Modeling and Control, Robotic Systems, Control of Uncertain Systems
Abstract: This paper presents a Nonlinear Model Predictive Control-based Reinforcement Learning (NMPC-based RL) framework for robot manipulators. The controller is developed to address the motion planning problem for robot manipulators in the presence of obstacles. The proposed control scheme includes a parametrized NMPC structure used as an approximator for the RL framework's value function and action-value function. In the NMPC structure, the cost function, system constraints, and the manipulator’s model are parameterized. The Q-Learning algorithm based on the Temporal Difference method adjusts the parameters of the NMPC to increase the closed-loop performance of the whole control scheme. The controller has been applied to a 6-degrees-of-freedom (DoF) model of a robot manipulator, aimed at moving its end-effector to reach the desired pose when static obstacles are in the robot's workspace. Numerical simulations demonstrate that the proposed controller can effectively control the end-effector's pose in such a way as to avoid any collisions between the manipulator and the obstacles. It is shown that the learning capability of the proposed NMPC-based RL framework can enhance the efficiency of the control loop up to 21%.

Keywords— Robot manipulator, Model predictive control, Reinforcement learning, Obstacle avoidance, Motion planning

Keywords: Intelligence Interaction, Information Visualization, Human Factors
Abstract: The motion signals are generated for a simulator user based on the visual understanding of the environment using virtual reality. In this respect, a motion cueing algorithm (MCA) is employed to reproduce the motion signals based on the real driving motion scenarios. Advanced MCAs are required to predict precise driving motion scenarios. Nonetheless, investigations on effective methods for predicting the driving motion scenarios accurately are limited. Current state-of-the-art studies mainly focus on the averaged motion signals from several simulator users pertaining to a specific map or from feedforward neural network and non-linear autoregressive. The existing methods are unable to yield precise predictions of the driving scenarios. In this research, the echo state network and long short-term memory models are employed for the first time in MCA to forecast the driving motion signals. Our evaluation proves the efficiency of our proposed methods in comparison with existing methods.

Keywords: Human-Machine Interface, Information Systems for Design, Human-Machine Cooperation and Systems
Abstract: To reduce the number of traffic fatalities, research and development of safety systems is currently under way. However, drivers are still required to judge traffic situations and operate vehicles safely. Therefore, the aim of this study is to clarify the effect of pedestrian information provided by a human–machine interface (HMI) on driving characteristics under different traffic conditions when vehicles turn right at intersections and drive straight where a pedestrian is crossing. The system is designed to present information on pedestrians to the drivers through the HMI. The experimental results reveal that the presentation of pedestrian information by the HMI increases the time to collision (TTC) and decreases closest distance to the pedestrian at intersections and on straight roads. Thus, it is effective for drivers to ensure safety. In contrast, results reveal that the presentation by the HMI has no effect on increasing the TTC or decreasing the closest distance to the pedestrian in the two types of driving situations; one case where the ego vehicle makes a right turn at the intersection just after the oncoming vehicle turns left, and another case where the ego vehicle goes straight with good visibility. Therefore, for a more effective HMI, a suitable design or timing of pedestrian information should be considered, depending on various traffic environments.

Keywords: Human-Computer Interaction
Abstract: Graph convolutional networks (GCNs) are widely used in skeleton-based action recognition and have achieved excellent results. However, it is evident that the convolution operation can lead to losing some original input information. The incomplete utilisation of original input data limits GCNs’ ability to obtain the skeleton’s correlation. This paper proposes a graph instinctive attention convolutional network (GIAN) to solve this problem. In particular, it contains an instinctive attention module that uses self-attention to obtain the correlation within the original input skeleton. Then, parameter attention is used to further refine the relationship between different skeleton joints. Experimental results on publicly available datasets demonstrate that the GIAN outperforms most of the state-of-the-art algorithms.

Keywords: Human-Computer Interaction, Intelligence Interaction, Augmented Cognition
Abstract: At present, there are about 70 million deaf people using sign language in the world, but for most normal people, it is difficult to understand the meaning of the sign language expression. Therefore, it is of great importance to explore the ways of recognising the sign language. In this paper, we propose a dynamic sign language recognition method based on the modified long-short-term memory (LSTM) model. Firstly, we use Leap Motion to collect the features of Chinese Sign Language (CSL). LSTM has a good effect in processing time series data, but the parameters of its hidden layer are shared, making it important information lost when dealing with long time series. The attention mechanism can give different attention weights to different features according to the correlation between the input data and output data, so as to enhance the model’s attention to key information. Therefore, we combine LSTM with an attention mechanism for dynamic sign language recognition. Experimental results show that the recognition accuracy of the modified LSTM model is 99.55%, which is higher than that of LSTM model. Finally, we developed a sign language human-computer interaction system, which verifies the real-time performance and effectiveness of the method proposed in this paper.

Keywords: Human-centered Learning, Medical Informatics, Assistive Technology
Abstract: This paper explores social capability properties via eye-hand coordination with a case study. In order to investigate social capability properties, an ICF based protocol is designed and experiments are conducted in an customized platform for investigating social capability for children with autism spectrum disorder. A set of eye-hand coordination based properties are extracted to represent individual social capability in terms of eye-hand coordination. A set of metric are also defined to measure the coordination behaviour such as engagement. It is evident that the eye-hand coordination based properties have huge potential for analyzing social capability, pave the way for machine-assisted intervention for ASD children.

Keywords: BMI Emerging Applications, Other Neurotechnology and Brain-Related Topics
Abstract: Brain-computer interfaces (BCIs) could enable persons with cervical spinal cord injury (SCI) to intuitively control assistive motor devices for regaining lost grasping function. Previous studies, mostly performed in non-disabled persons, have already shown that complex upper limb movements can be decoded from the low-frequency time domain of the electroencephalogram (EEG). In this work, we attempted to translate these results to persons with cervical SCI and investigated whether executed reach and attempted grasp actions could be decoded from their EEG signals. For this, we chose three different reach-and-grasp actions, two unimanual and one bimanual, towards objects of daily life. During participants’ self-initiated, executed reach and attempted grasp actions, we recorded EEG using mobile, water-based electrodes. We measured two participants with subacute cervical SCI who had preserved shoulder movements and elbow flexion but no wrist and hand functions. Both repeated the session three times. We also recorded the EEG of 10 non-disabled persons performing the same tasks (control group). We extracted and analyzed movement-related cortical potentials (MRCPs) from the EEG’s low-frequency time domain. Consecutively, we assessed the decoding capabilities of two linear (shrinkage based linear discriminant analysis (sLDA), linear support vector machine (SVM)) and two non-linear (Random Forests (RF), naıve Bayes (NBC)) classification models for the discrimination of the grasp actions. We could show that sLDA, SVM, and Random Forest yielded comparable classification results on average with 63.4% ± SD 9% for participants with SCI and 69.7% ± SD 9% for the control group (chance level 29.3%). Our results indicate that it is feasible to decode executed reach and attempted grasp actions from MRCPs of persons with subacute cervical SCI. Future measurements will provide additional data to assess the generalizability of our results in a larger group of people with cervical SCI.

Keywords: Active BMIs
Abstract: Motor imagery (MI) based brain-computer interfaces (BCIs) have been studied as applications for the improving rehabilitation and recovery, as well as augmenting existing function. MI BCI systems typically provide feedback in an egocentric rather than an allocentric reference frame. The goal of this study is to see if presenting stimuli in an allocentric reference frame is comparable to presenting egocentric stimuli. We used dynamic visual stimuli in egocentric and allocentric reference frames to induce motor imagery in a virtual reality (VR) environment. Eight participants participated in this study, in which they imagined grasping actions with their left and right hands while observing egocentric or allocentric stimuli. The allocentric and egocentric reference frame tasks had comparable inter-rater agreement and precision, indicating that allocentric visual feedback is as effective as egocentric one for MI BCI.

Keywords: Active BMIs
Abstract: Active BCIs often rely on the performance of various imagery tasks that can be detected and decoded as commands for operating an external device. An increase in the number of these commands would significantly improve the functionality of the system and, as a result, quality of life for the user. This study focuses on augmenting the practicality and performance of such systems by investigating the mental task of singing imagery. Singing imagery is the simple act of imagining singing a song in your head and is a common experience for most people. However, despite its intuitive and straightforward nature, the potential of singing imagery for improving the performance of active BCIs or increasing their number of commands has not been fully investigated.

In this study, along with various binary analyses, singing imagery is combined with a set of conventional tasks in BCI design (i.e., the motor imagery of large body parts like hands, feet, and tongue) for 4-class, 5-class (adding a “rest” state), and ultimately 6-class scenarios to evaluate the possibilities of enhancing the active BCI systems. Incorporating a singing imagery task in the 4 and 5-class combinations yielded up to 6.6% and 6.7% improvements in classification accuracy for the two cases, respectively. Furthermore, for the 6-class scenario, accuracies as high as 49.8%, which is well above the chance level of 16.7%, were achieved.

The preliminary results, obtained from five participants, suggested that singing imagery could be a viable option for improving the classification accuracy or increasing the number of commands.

Keywords: Other Neurotechnology and Brain-Related Topics, Active BMIs
Abstract: Brain-computer interfaces (BCIs), such as the P300 speller, can provide a means of communication for individuals with severe neuromuscular limitations. BCIs interpret electroencephalography (EEG) signals in order to translate embedded information about a user's intent into executable commands to control external devices. However, EEG signals are inherently noisy and nonstationary, posing a challenge to extended BCI use. Conventionally, a BCI classifier is trained via supervised learning in an offline calibration session; once trained, the classifier is deployed for online use and is not updated. As the statistics of a BCI user’s EEG data change over time, the performance of a static classifier may decline with extended use. It is therefore desirable to automatically adapt the classifier to current data statistics without requiring offline recalibration. In an existing semi-supervised learning approach, the classifier is trained on labeled EEG data and is then updated using incoming unlabeled EEG data and classifier-predicted labels. To reduce the risk of learning from incorrect predictions, a threshold is imposed to exclude unlabeled data with low-confidence label predictions from the expanded training set when retraining the adaptive classifier. In this work, we propose the use of a language model for spelling error correction and disambiguation to provide information about label correctness during semi-supervised learning. Results from simulations with multi-session P300 speller user EEG data demonstrate that our language-guided semi-supervised approach significantly improves spelling accuracy relative to conventional BCI calibration and threshold-based semi-supervised learning.

Keywords: Human-Machine Interface, Human-Computer Interaction
Abstract: This paper presents a new feature extraction method for Electroencephalogram (EEG)-based motor imagery (MI) classification. Current researches mostly classify different MIs by detecting the event-related desynchronization (ERD) phenomenon from the EEG signals. Due to the poor spatial resolution of the MI-EEG signals, the cortical area (source) activating the MI cannot be located accurately with the EEG (sensor) signal, which might degrade the classification accuracy. This study adopts the EEG source imaging (ESI) technique to estimate the cortical area where source ERD happens from the EEG signal. An improved ESI method based on the linearly constrained minimum variance (LCMV) algorithm, in which an average LCMV filter and an average baseline covariance are constructed for the ESI, is proposed to locate the activated cortical area from the noisy EEG signals. The source ERD features are then extracted. Analytical results show that, for subjects with obvious average source ERD phenomenon, their activated cortical area in a single-trial MI can be well located. MI classification results also support the feasibility of the proposed method for MI-EEG signal processing.

Keywords: Distributed Intelligent Systems, Smart Sensor Networks, System Modeling and Control
Abstract: The goal of this research is to improve the wireless sensor network data transmission method to save energy consumption of sensor nodes and improve the overall network life cycle. This study uses the ant colony algorithm task allocation control strategy, the execution capability, energy consumption and lifespan of each central node to realize the real-time task allocation control. This study proposes to use an improved hybrid ant colony algorithm to find the shortest path of the cluster and transmit it to the confluence node, so as to avoid unnecessary energy consumption caused by long-distance transmission. Through the independent cooperation of each node and the completion of data processing, the analysis results are finally obtained through information fusion and comprehensive decision-making. The research experiments show that the peer-to-peer wireless network structure in this study not only reduces the burden of communication, calculation and energy consumption of a single central node, but also helps to improve the life and stability of the wireless sensing network.

Keywords: Distributed Intelligent Systems, System Modeling and Control, Large-Scale System of Systems
Abstract: In this study, we propose a modified predictive direct torque control (PDTC) application-specific integrated circuit (ASIC), comprising a neural network (NN) proportional integral derivative (PID) controller, speed-sensorless control, fuzzy error controller, and seven-stage hysteresis controller, to alleviate the ripple problem induced by limited vector voltages and slow speed response in conventional direct torque control. Both flux and torque errors pass through the modified discrete multiple vector voltage switch table to obtain the required vector voltages, and the proposed NN PID controller is used to convert the speed error into a torque command. Notably, the motor speed is evaluated from the magnetic flux, which is calculated using two-phase currents and voltages. The speed-sensorless control not only accelerates the feedback control but also rotates more stably. The NN PID controller generates a torque command according to the speed error, which is obtained by subtracting the estimated predictive speed from the actual speed. The advantages of the proposed system are that it reduces the flux and torque ripples and increases the control stability by filtering out the external interferences. The Verilog hardware description language is used to implement the proposed PDTC ASIC system, and a field-programmable gate array development board is used to verify the designed functions.

Keywords: Intelligent Transportation Systems, Robotic Systems, Modeling of Autonomous Systems
Abstract: The development of monocular visual simultaneo-us localization and mapping (VSLAM) has slowly begun in recent years. At present, the sensors used for VSLAM include monocular, binocular, or depth. For visual mapping, two problems will be encountered and the mapping cannot be performed. One is when there are not enough feature points, the camera's pose at the next moment cannot be estimated, such as a wall. The other is the VSLAM of dynamic environment changes may not be recognized as the same object because the feature matching needs to be coded through its surrounding environment, so it is easy to lose track when encountering changes in light. Compared with binocular and depth, monocular vision lacks depth information, but because it is cheap and easy to install, it needs to be used by multiple people. The current research proposes adding other information, including the camera height, the scene of the reference object and depth estimation by learning methods. The study uses the OpenVSLAM architecture to estimate the camera height scale, and proposes the mechanism be based on the change in the average scale of the first five key frames, with the average scale being updated at the same time to correct the current scale. Through this method, the drastic changes in scale are corrected, and the accuracy of trajectory positioning is improved. We also evaluate our proposed method on a real KITTI Dataset and demonstrate the proposed algorithm is effective and feasible for monocular visual SLAM.

Keywords: Distributed Intelligent Systems
Abstract: DNA methylation plays an important role in the regulation of gene expression, and aberrant changes in epigenetic regulation can be detected in cancer development or progress. In this study, genome-wide DNA methylation profiles and electronic medical records were combined to identify effective DNA methylation biomarkers for colorectal cancer (CRC). Statistical analytics and deep learning approaches were integrated to explore associated novel biomarkers. These identified biomarkers could facilitate accurate diagnosis of CRC and monitor disease progression for a testing subject, and provide suggestions for appropriate medical treatments. Firstly, DNA methylation profiles were analyzed through a standard pipeline to discover significantly differential methylation biomarkers as primary biomarkers. Incorporating Taiwan’s National Health Insurance Research Database (NHIRD), associated comorbidities of CRC were identified and associated disease genes were considered as secondary biomarkers. The intersection of primary and secondary biomarkers was performed to obtain CRC-relevant biomarker candidates, and gene ontology annotations were used to calculate candidate biomarker-to-biomarker distances. Based on the formulated gene-pair distance matrix for all candidate biomarkers, we applied clustering algorithms to categorize candidate biomarkers into different functional groups. Corresponding coefficients for each biomarker within a functional cluster were ranked through an attention-based recurrent neural network. The weighting coefficient for each biomarker was further taken as a reference for designing a practical testing toolkit. Finally, we obtained 11 important biomarkers from 5 functional clustered groups which were established by the 141 biomarker candidates screened from 450k DNA methylation probes.

Keywords: Distributed Intelligent Systems, Communications, Smart Sensor Networks
Abstract: This paper presents an Ethernet packet transformation and transmission architecture between Modbus transmission control protocol (Modbus/TCP) and universal serial bus (USB) 3.0 developed with a field-programmable gate array (FPGA) development board. The proposed architecture is used to complete packet transformation and transmission between Ethernet and USB 3.0 for application in plant automation. The Ethernet receiver receives and analyzes Modbus/TCP packets and sends the source address, destination address, IP header, and Modbus/TCP header to the register to verify the correctness of the packet. The Modbus/TCP packet is stored in static random access memory and awaits access by a USB 3.0 module. An FPGA development board (Intel DE10-Standard) is used for functional verification. The measured results show that the latency, throughput, and dynamic power are 18.845 μs, 747.45 Mbps, and 142.17 mW, respectively, at a voltage of 1.8 V and operating frequency of 125 MHz.

Keywords: Image Processing and Pattern Recognition, Deep Learning, Neural Networks and their Applications
Abstract: Deep subspace clustering methods have attracted extensive attention due to the great improvement in both representation ability and precision of non-linear data. However, the rich information which is contained in the self-expression matrix is unexplored since existing approaches use the self-expression matrix only as a tool for learning inter-sample relationships and clustering. In addition, such models treat outlier and noise points equally with other points, which inevitably degrades the clustering performance. To overcome these issues, we develop an progressive deep subspace clustering approach by extracting delayed fitting probabilities from the module and then use the probabilities to defer the fitting of unreliable points. Specifically, we calculate the probabilities that each sample lies in each subspace based on the results of the self-expression matrix and spectral clustering, and then estimate the reliability of the cluster assignment of each sample as delayed fitting probability to reweight the loss of each sample. Experiments on five benchmark datasets validate the effectiveness of the proposed method.

Keywords: Deep Learning, Image Processing and Pattern Recognition, Multimedia Computation
Abstract: Image-text matching plays a crucial role in connecting vision and language. The details of the objects in the image, the positional relationship, and the correspondence between the background and the text description are the keys to image-text matching. Previous studies either only extract the salient objects of the image, or only pay attention to the location of the object, ignoring the detailed features and background features of the object, and the extraction of the overall semantic information of the image is not comprehensive enough. Accordingly, this paper proposes a model based on Bert and Self-Attention Mechanism (BSAM), we segment the image area, use the self-attention mechanism to enhance the weight of the key area, pay attention to each object and their detailed features and background features, the image regions are mapped into original region features and new features with other region relationships, and the global information of the image is inferred based on the relationship between each region and background features. The text extracts word features and new features with other word relationships through the Bert model. We propose the Cross-Attention and Similarity -Attention Filtering (CA-SAF) module to align all relevant image regions and words, enhance matching pairs with high weights, and filter matching pairs with lower weights. Extensive experiments on two datasets, Flickr30K and MS COCO, show that the BSAM model significantly outperforms state-of-the-art methods.

Keywords: Machine Learning, Heuristic Algorithms, Image Processing and Pattern Recognition
Abstract: In supervised learning, the amount of work required to assign labels to data becomes enormous as the input data increases. To reduce this work, we propose a method for label estimation using only a small amount of labeled data. The SVM separation hyperplane is moved by iterating label assignment and classification, and the results are evaluated using a new cluster evaluation criterion. This criterion is based on the Fisher criterion used for linearly separable data and applied to nonlinear data. Clustering and semi-supervised learning, which are related to the proposed method, are also explained. We show that the proposed method is effective for linearly inseparable data through experiments. In addition, we compared the performance of the proposed method with clustering and semi-supervised learning.

Keywords: Image Processing and Pattern Recognition, Deep Learning, Neural Networks and their Applications
Abstract: In recent years, Generative Adversarial Networks (GANs) have gained attention in image synthesis mapping from the latent space onto image space. Trained latent space carries the visual semantics for generated images. Past studies observed that arithmetic operation and linear interpolation in latent space could change the visible facial attributes, such as beards and glasses, in image space. In this work, the visual concepts in the latent space are observed, allowing to change the emotion attribute per facial expressions in the image space. We observed interpolation of a sample while disentangling the emotional attributes to edit the emotion-related facial expressions in the synthesized images. For the experiment, the Deep Convolution Generative Adversarial Networks (DCGANs) are utilized for image synthesis, and Extended Cohn Kanade (CK+) facial expression dataset is applied as the input. Our results showed that manipulating the latent space of the well-trained GANs can edit the emotional aspects of the image space. Moreover, editing facial expressions in the latent space is helpful for the recognition task to improve accuracy. Empirical results showed that the facial expressions classifier improved its performance in the recognition sadness class from 20% to 80% on the imbalance dataset.

Keywords: Image Processing and Pattern Recognition, Media Computing, Deep Learning
Abstract: Recently, state-of-the-art unsupervised re-ID methods train the neural network by calculating cluster-level or instance-level contrastive loss for learning discriminative features with unlabeled data. However, due to the divergence of the individual cluster, the previous methods did not fully utilize inherent feature of clustered samples for contrastive learning, which merged an unreliable instance into a wrong cluster by simply using cluster centroid or hard instance. To solve this issue, we propose a novel Cluster-aware Diversity Samples Mining (CDSM) framework based on the compactness and independence of individual cluster to generate diverse samples for updating memory dictionary of each cluster, so as to reduce the effect of noisy labels and improve the robustness of the model performance. Significantly, the proposed Cluster-aware Diversity Samples Mining method gradually creates more reliable clusters to generate more robust pseudo labels by refining the memory which is of central importance to our outstanding performance. Extensive experiments demonstrate that the proposed CDSM framework achieves performances of 85.6%, 73.7% and 31.0% mAP on Market1501, DukeMTMC-reID and MSMT17, respectively. Code is available at https://github.com/colinzhaoxp/CDSM.

Keywords: Human Performance Modeling, Assistive Technology
Abstract: With the advent of the information society in recent years, ideas are becoming more and more valuable, and divergent thinking is becoming more and more important. Although there are methods to present visual stimuli to improve divergent thinking, these methods mainly present visual stimuli to the central visual field. The purpose of this study is to test hypotheses about changes in the number and variety of divergent thinking ideas under conditions in which visual stimuli, called geometric figure animations, are presented in the peripheral visual field. The evaluation experiment was conducted with the participation of 32 university students. We asked them to perform the Alternative Uses Test (AUT) under two conditions: presenting and non-presenting geometric figure animations in the peripheral visual field. The performance of divergent thinking was evaluated by measuring the number of responses, fluency, and flexibility of the AUT. The scores were standardized because the AUT differed in difficulty depending on each AUT topic. The results showed that all scores were higher in the presenting condition than in the non-presenting condition, and the effect size of the mean difference in flexibility was moderate. On the other hand, according to the results of the questionnaire, most of the participants did not feel that the geometric figure animations provided any cues for divergent thinking. This suggests that the flexibility of divergent thinking might be improved unconsciously by geometric figure animations presented in the peripheral visual field.

Keywords: Human Factors, Human Performance Modeling, Human-Machine Interface
Abstract: In this paper, focusing on the relationship between tendon tension and force applied to the fingertip, we develop a ring-type sensor that estimate fingertip force from tendon tension. To establish the measurement principle, mechanical models were constructed in three steps for sensor structure, deformation of finger and sensor, and finger kinematics. By integrating those mechanical relationships, we derived a relation expression between ring-type sensor output and fingertip force. In addition, we conducted experiments in which the developed ring-type fingertip force sensor was attached to the index finger and force was applied to the fingertip. As the results, it was confirmed that the fingertip force could be estimated by using the derived relational expressions both when the finger was extended and when it was bent. The results showed the validity of the mechanical models and the effectiveness of the sensor. These results provide important insights that will be useful for future sensor design and applications.

Keywords: Brain-based Information Communications
Abstract: Functional near-infrared spectroscopy (fNIRS) is a brain imaging method introduced relatively recently, which is promising to implement brain-computer interfaces. However, there is still a lack of research on fNIRS signal classification, particularly that focusing on improved machine learning techniques for non-motor tasks. In this paper, we propose a novel deep learning method using brain connectivity for resting-state fNIRS signal classification. Our method is based on the powerful modeling capability of the convolutional neural network that learns the brain connectivity patterns residing in the fNIRS signal. In particular, we present a new data augmentation method that can overcome the scarcity of fNIRS data. Experimental results of subject-independent classification of flourishing levels demonstrate the superiority of our approach to conventional approaches. It is also shown that the data augmentation strategy is effective for improving classification performance.

Keywords: Human-Machine Interface, Brain-based Information Communications, Intelligence Interaction
Abstract: Recent development in implantable devices for electrical brain stimulation includes sensing and embedded computing capabilities that enable adaptive stimulation strategies. Applications include stimulation triggered by pathologic brain activity and endogenous rhythms, such as circadian rhythms. We developed and tested a system that integrates an electrical brain stimulation & sensing implantable device with embedded computing and uses a distributed system with commercial electronics, smartphone and smartwatch for patient annotations, extensive behavioral testing, and adaptive stimulation in subjects in their natural environments. The system enables precise time synchronization of the external components with the brain stimulating device and is coupled with automated analysis of continuous streaming electrophysiology synchronized with patient reports. The system leverages a real-time bi-directional interface between devices and patients with epilepsy living in their natural environment.

Keywords: Human-Machine Interface, Augmented Cognition, Intelligence Interaction
Abstract: Electrical stimulation of deep brain structures is an established therapy for drug-resistant focal epilepsy. The emerging implantable neural sensing and stimulating (INSS) technology enables simultaneous delivery of chronic deep brain stimulation (DBS) and recording of electrical brain activity from deep brain structures while patients live in their home environment. Long-term intracranial electroencephalography (iEEG) iEEG signals recorded by INSS devices represent an opportunity to investigate brain neurophysiology and how DBS affects neural circuits. However, novel algorithms and data processing pipelines need to be developed to facilitate research of these long-term iEEG signals. Early-stage analytical infrastructure development for INSS applications can be limited by lacking iEEG data that might not always be available. Here, we investigate the feasibility of utilizing the Deep Generative Adversarial Network (DCGAN) for synthetic iEEG data generation. We trained DCGAN using 3-second iEEG segments and validated synthetic iEEG usability by training a classification model, using synthetic iEEG only and providing a good classification performance on unseen real iEEG with an F1 score 0.849. Subsequently, we demonstrated the feasibility of utilizing the synthetic iEEG in the INSS application development by training a deep learning network for DBS artifact removal using synthetic data only and demonstrated the performance on real iEEG signals. The presented strategy of on- demand generating synthetic iEEG will benefit early-stage algorithm development for INSS applications.

Keywords: Decision Support Systems, Enterprise Information Systems, Discrete Event Systems
Abstract: Surgical center scheduling is challenging to schedule physical spaces to reduce costs and increase productivity. This study presents a framework called PM4SOS (Process Mining for Simulation, Optimization, and Scheduling) that facilitates the generation of an operating room schedule in an automated way and integrates, reducing cognitive overload and waste of time. This framework allows the evaluation of restrictions to get the best performance of surgical schedules. PM4SOS combines process mining with data from event logs, generation of the automated simulation model, and case-based reasoning (CBR) to analyze management indicators, allowing significant gains in optimization of the scheduling of surgical centers. The results with PM4SOS help decision-making in hospital environments to better use physical resources and human resources, such as reducing waiting time, optimizing surgery execution time, and resource capacity management.

Keywords: Decision Support Systems, Consumer and Industrial Applications
Abstract: When investors invest in the stock market, the first decision is invested in which stock market and how to decide on the invested stock. The high return stock is accompanied by high risk. Constructing a portfolio can help to diversify the risk in the investment. This paper proposes a decision support system that can help select stocks and construct a portfolio with high return and low risk precisely in the Singapore stock market, which is one of the top performances in Asia. The proposed system uses the novel assessment indicator trend ratio to simultaneously consider the portfolio return and risk and then uses the global-best guided quantum-inspired tabu search algorithm with quantum-NOT gate (GNQTS) to search for the near-optimal solution. The decision-making system also includes 13 sliding windows to retain the freshest data and identify the suitable investment period. According to the experimental results, the system consisting of evolutionary computation can perform robust and rational decisions to efficiently construct a stable uptrend portfolio with the highest trend ratio in the Singapore stock market.

Keywords: Decision Support Systems, System Modeling and Control, Fault Monitoring and Diagnosis
Abstract: This paper contributes to the understanding of orthogonalization approaches widely used in system identification, signal processing, machine learning, and automation. Generalized orthogonalization is proposed that provides a unified, alternative formulation to Gram-Schmidt orthogonalization, Singular Value Decomposition, and Principal Component Analysis over finite-dimensional Euclidean spaces. The proposed approach puts SVD and PCA into the perspective of operations research, providing (a) additional insights into their distinctive features and (b) foundations for the extensions to inner product and probability spaces.

Keywords: Decision Support Systems, Intelligent Transportation Systems
Abstract: The reinforcement learning (RL) with deep neural network, as a data-driven approach, is promising for adaptive traffic signal control (ATSC) in traffic scenarios. The majority of the existing studies focus on designing efficient agents and policy optimization for ATSC, but neglect to observe more detailed states of the environment. In this paper, an adaptive traffic signal control strategy, named as A2C_RTQL, is proposed for scheduling the traffic signal in an intersection, by combining the real-time lane-based queue lengths with deep RL agent. First, the Lighthill-Whitham-Richards (LWR) shockwave theory is employed for obtaining the real-time queue lengths in each lane. After that, by defining the obtained queue lengths as the inputs, A2C_RTQL strategy is designed for traffic signal control based on the advanced actor-critic (A2C) agent, where the lanes are divided into multiple parallel environments based on the phases of traffic signal. Simulation results demonstrate the optimality and efficiency of the proposed strategy compared with other methods in SUMO under simulated peak-hour traffic dynamics.

Keywords: Neural Networks and their Applications
Abstract: This papers describes a learning algorithm for growing neural gas to construct a topology-preserving map from a 3D point cloud whose topology can change dynamically. Growing Neural Gas with Utility Factor (GNG-U) has been presented as a method for learning the topology of a 3D space environment and applying it to non-stationary or dynamic data distribution. However, when a node is added to an existing network after several errors with sampling data have accumulated, it is difficult for a standard GNG-U to considerably boost learning speed. As a result, we propose a revolutionary growth strategy that dramatically accelerates learning and convergence. This method immediately adds a sample of data as a new node to an existing network based on the likelihood of node addition estimated by the distance to the third closest node and the first and second closest nodes at maximum. Experiment findings show that the proposed algorithm's network can quickly adapt to represent the topology of non-stationary input distributions.

Keywords: Infrastructure Systems and Services, Intelligent Transportation Systems, Smart Buildings, Smart Cities and Infrastructures
Abstract: Autonomous driving poses high demands on computing and communication resources. Vehicular edge computing is presented to offload real-time computing tasks from connected and automated vehicles (CAVs) to high-performance edge servers. However, it brings additional communication overhead due to limited bandwidth, and increases delay of tasks. To solve it, this work first proposes an offloading architecture including multiple CAVs, roadside units and cloud. We minimize the total cost of a hybrid system by jointly considering task offloading ratios, and allocation of communication and computing resources. Furthermore, a mixed integer non-linear program is formulated and solved by a novel meta-heuristic algorithm called Self-adaptive Gray Wolf Optimizer with Genetic Operations (SGWOGO). SGWOGO achieves joint optimization of computation offloading among CAVs, roadside units, and cloud, and allocation of their resources. Finally, real-life data-driven simulation results demonstrate that SGWOGO achieves lower cost in fewer iterations compared with its several state-of-the-art peers.

Keywords: Quality and Reliability Engineering, Electric Vehicles and Electric Vehicle Supply Equipment
Abstract: Software trustworthiness is a composite reflection of software quality and dependability attributes that are defined in industrial standards (e.g., ISO 25010), indicating a software system is constructed and operated as expected. Trustworthiness evaluation has become increasingly vital for software production and its permission being used in industry. However, trustworthiness evaluation is challenging due to the absence of comprehensive models, systematic methods, and efficient tools. We present iTrustEval a framework for software trustworthiness evaluation with an intelligent analytic hierarchy process (AHP)-based method. In iTrustEval an extensible trustworthiness model enabling on-demand integration with industrial trustworthy standards (such as ISO 25010 and Automotive SPICE in the current model) is proposed; an AHP based method is designed for the bottom-up measuring data fusion (where a hybrid missing-value recommendation engine is developed using both temporal-attenuation-mechanism based history data recommendation and matrix factorisation-based recommender system); and a prototypical tool has been developed. The applicability of iTrustEval is validated through a case study, and the results show it is sound in efficiency and effectiveness.

Keywords: Intelligent Transportation Systems, Modeling of Autonomous Systems
Abstract: Safety issues related to autonomous vehicles are of great concern both in the academy and industry and the braking system performance is a crucial research field. In this work, an autonomous braking system based on deep reinforcement learning is proposed, employing an intelligent agent trained in city scenarios to manage both pedestrians’ safety and passengers’ comfort. The agent is modelled via the deep deterministic policy gradient algorithm in a simulation environment and its performance is tested showing good results for guaranteeing both pedestrians’ safety and passengers’ comfort.

Keywords: Smart Sensor Networks, Communications, Distributed Intelligent Systems
Abstract: 在移动自组织可穿戴无线传感器网络中，使用通信协议来保持数据的实时性、可靠性和同步准确性。这种移动自组织网络采用点对点的星形网络结构，分为三个层次：传感器节点、汇聚节点-协调器、移动终端。为了解决多惯性传感器节点的数据拥塞和低同步效率问题，设计并实现了一种基于FreeRTOS的自主通信协议。在此基础上，通过基于TPSN的改进时间同步算法，将多惯性传感器节点的同步时延控制在5ms以内，进一步保证了主协调器与从节点之间的数据处理频率相同。

Keywords: Medical Mechatronics, Bio-mechatronics and Bio-robotics Systems
Abstract: Acoustoelectric impedance tomography (AET) is a new non-invasive medical imaging procedure used to map the electrical properties of biological tissues with higher spatial resolution than traditional electrical impedance tomography (EIT). It exploits the acoustoelectric effect where modulated ultrasonic pressure changes the local tissue conductivity. This provides additional information to reconstruct a tomographic image, and has a stabilizing effect on an otherwise highly unstable inverse problem.

In this paper, a novel approach to solving the AET inverse problem for image reconstruction is proposed. In the algorithm, the acoustoelectric effect is assumed to create small perturbations in the local resistance of the medium under observation. A lumped model consisting of a finite mesh of resistors approximates the medium under observation, through which boundary voltage differences between the excited and unexcited medium are calculated. A variation of the Modified Newton Raphson (MNR) algorithm is then proposed, where each pattern in the algorithm is created from small perturbations of the tissue conductivity. A total of eight simulation scenarios are evaluated, where the conductivity perturbations are in the order of 1%, 2.5% to 5% of the nominal tissue conductivity. The algorithm can successfully reconstruct the images in the presence of random noise. The obtained images are compared against traditional EIT where the percentage error is calculated for each simulated tomographic image. The simulation results indicate that the proposed approach is superior to traditional EIT as it constructs more distinct and high contrasting images with less percentage error.

Keywords: Hybrid Models of Neural Networks, Fuzzy Systems, and Evolutionary Computing, Image Processing and Pattern Recognition, Deep Learning
Abstract: Analyzing the dynamics of neural activity patterns using an electrophysiological approach is important for un- derstanding the basis of information processing in a brain. In this study, we attempted to extract units of neuronal activity representing certain information and to discriminate activity pattern evoked by electrical stimulus and spontaneous activity without stimulation. We applied X-means clustering to the triplet of“2D-spacial coordinates and timestamps”of neuronal electrical spike bursts and the extracted single spatiotemporal neuronal activity pattern was converted into a standardized 8×8 2D-spatial pattern map. Most 2D-spatial patterns emerged repeatedly during recording time, corresponding to a repre- sentative motif. Then we gathered similar 2D-spatial patterns (primary clusters) into one ”pattern repertoire” and attempted to classify these pattern repertoires depending on the stimulus inducing the pattern repertoire by VGG16 deep convolutional neural networks (deep CNN). For that, we prepared two types of 224×224 images by converting the 2D-spatial pattern maps, the spatial information priority neural activity pattern image (SIP-NAP image) and the time information priority neural activity pattern image (TIP-NAP image). As a result, over 80% of high discrimination accuracy was obtained for both types of VGG16 input images, especially 99.7% for TIP-NAP image. It suggested that spatiotemporal features contributing to the discrimination between spontaneous and evoked response activities were extracted by 3-D-crustering. In addition, it is shown that the spatial pattern of neuroelectric activity is able to be discriminated with high accuracy by combining transfer learning and SIP-NAP / TIP-NAP image translation method, even with a small amount of training data.

Keywords: Deep Learning, Neural Networks and their Applications, Application of Artificial Intelligence
Abstract: 超声成像广泛应用于胎儿的产前诊断中，以监测胎儿发育，具有重要的临床意义。 产前诊断的主要目的是测量胎儿的一些生物学参数，如腹围（AC）、股骨长度（FL）、冠臀长度（CRL）、双翅直径（BPD）、头围（HC）等。 然而，胎儿超声图像中的伪影，噪声和颅骨丢失对胎儿HC的自动测量提出了一些挑战。 为了提高胎儿HC自动测量的准确性，提出具有后处理的FCS-Unet。 首先，通过Faster R-CNN进行胎头定位，找出胎头定位箱的坐标。 然后对于胎儿头围分割，FPN用于代替Unet的编码器。 在跳过连接中，通道和空间模块通过求和运算进行合并。 在后处理部分，生成的二进制图像中头部定位框外的白色像素被归零并变为黑色。 最后，采用最小二乘椭圆拟合法计算HC，ô

Keywords: Deep Learning, Machine Learning, Transfer Learning
Abstract: Currently, deep learning is widely used in the field of medicine, which in turn includes radiology. This paper considers the problem of the classification of X-ray images and the lack of images of specific classes. The classes included COVID-19 and Normal X-ray scans. To solve the problems, we propose few-shot learning that is based on different Residual Convolutional Neural Network models with different complexities. The method is designed for the datasets that have small amount of samples of a specific class and a larger amount of instances of another class. The utilization of few-shot learning can solve the issues of the balance of X-ray datasets. The Residual Convolutional Neural Network models we used are as follows: ResNet-50, ResNet-101, and ResNet-152. textcolor{black}{ The architectures had been used to extract the features from the images that were used later.} The latter model has the highest complexity, while the former has the lowest complexity, respectively. The obtained results include the highest accuracy of 97.7% for 10 shots of COVID-19 positive X-ray images. The accuracy was achieved using ResNet-101 model. The highest result for ResNet-152 model was 95.6 %. However, on average, the model achieved the highest accuracy. ResNet-50 model provided the least accurate results, however, it is less complex which provides faster performance. One can also notice that with the higher number of COVID-19 positive shots that were used for training, the accuracy also gets higher. To provide transparency to our solution, we furthermore created t-distributed stochastic neighbor embedding visualization. This showed us that the system could separate the two classes into two distinct clusters. Overall, the results imply the efficiency of the solution that was proposed in the study.

Keywords: Image Processing and Pattern Recognition, Neural Networks and their Applications, Application of Artificial Intelligence
Abstract: Since vaccination started, the COVID-19 scenario has improved. On the other hand, although the number of deaths has significantly dropped, the number of new cases is still a concern. Thus, patient tracking and follow-up are essential tasks, and chest X-ray examination is the first-order tool. While several studies using CXR and computing have been developed, they did not translate into clinical applications yet. One of the reasons is the computational effort required to run huge deep learning models and its high cost to be adopted in community clinics. Therefore, this work proposes a lightweight (few computational resources needed), fast (training and inference time), and reasoned solution for automatic COVID-19 detection and assessment of its severity. Our method is based on extracting features by Binary Pattern of Phase Congruency (BPPC) in segmented CXR images. Radiomic features are extracted from the segmented CXR image, and an SVM-based selection process is used to build two models of a shallow Feed-Forward network. The results surpass previous studies, with an average accuracy for COVID-19 detection of 98.71%. For images without evidence of infection but with a positive PCR test, an accuracy of 94.74% is reached. In a second task, the severity level of COVID 19 is estimated with an AUC of 98.92%. This high performance helps improve the speed and accuracy of diagnosis and severity assessment of COVID19 infection, proving to be a viable option in transitioning from a research field to a clinical environment.

Keywords: Computational Intelligence, Machine Learning, Neural Networks and their Applications
Abstract: In recent engineering applications using deep learning, physics-informed neural network (PINN) is a new development as it can exploit the underlying physics of engineering systems. The novelty of PINN lies in the use of partial differential equations (PDE) for the loss function. Most PINNs are implemented using automatic differentiation (AD) for training the PDE loss functions. A lesser well-known study is the use of finite difference method (FDM) as an alternative. Unlike an AD based PINN, an immediate benefit of using a FDM based PINN is low implementation cost. In this paper, we propose the use of finite difference method for estimating the PDE loss functions in PINN. Our work is inspired by computational analysis in electromagnetic systems that traditionally solve Laplace's equation using successive over-relaxation. In the case of Laplace's equation, our PINN approach can be seen as taking the Laplacian filter response of the neural network output as the loss function. Thus, the implementation of PINN can be very simple. In our experiments, we tested PINN on Laplace's equation and Burger's equation. We showed that using FDM, PINN consistently outperforms non-PINN based deep learning. When comparing to AD based PINNs, we showed that our method is faster to compute as well as on par in terms of error reduction.

Keywords: Cloud, IoT, and Robotics Integration, Optimization and Self-Organization Approaches
Abstract: Digital twins present revolutionary potential in smart manufacturing and production. However, their application to distributed control systems is limited in literature. This leaves automation engineers wishing to apply the concept at a loss, as they must construct a digital twin from a conceptual level with little to no guidelines. To help the adoption of the digital twin concept, a general architecture that may be tuned to the needs of the physical application is required.

Within this paper, we propose an architecture for digital twins in IEC61499 based distributed control systems. Using this architecture, we construct the model based on known physics and sensor data to be used in the digital twin for simulation purposes. This demonstrates itself as an effective method for constructing the model of a digital twin for the purposes of dynamic simulation and control. With this base architecture in place, we may now work towards expanding the capability of the set-up to that of a full digital twin.

Keywords: Cybernetics for Informatics, Application of Artificial Intelligence, Computational Intelligence
Abstract: As more novel devices are integrated into the electricity grid due to the changes taking place in the energy system, ways of detecting deviations from the intended settings are needed. If misconfigurations of, for example, reactive power control curves of inverters go unnoticed, the safe and reliable operation of the power grid can no longer be ensured due to possible voltage violations or overloadings. Therefore, methods of detection of misconfigurations of said inverters using operational data at transformers are presented and compared. These methods include preprocessing by dimensionality reduction as well as detection by supervised learning approaches. The data used is of high reliability as it was collected in a lab setting reenacting typical and relevant grid operation situations. Furthermore, this data was recreated by simulation to validate the simulation data, which could also possibly be used for detection applications on a bigger scale. The results for both data sources were compared and conclusions drawn about applicability and usability for grid operators

Keywords: Machine Learning, Deep Learning, Application of Artificial Intelligence
Abstract: Indoor thermal comfort in smart buildings has a significant impact on the health and performance of occupants. Consequently, machine learning (ML) is increasingly used to solve challenges related to indoor thermal comfort. Temporal variability of thermal comfort perception is an important problem that regulates occupant well-being and energy consumption. However, in most ML-based thermal comfort studies, temporal aspects such as the time of day, circadian rhythm, and outdoor temperature are not considered. This work addresses these problems. It investigates the impact of circadian rhythm and outdoor temperature on the prediction accuracy and classification performance of ML models. The data is gathered through month-long field experiments carried out in 14 classrooms of 5 schools, involving 512 primary school students. Four thermal comfort metrics are considered as the outputs of Deep Neural Networks and Support Vector Machine models for the dataset. The effect of temporal variability on school children’s comfort is shown through a “time of day” analysis. Temporal variability in prediction accuracy is demonstrated (up to 80%). Furthermore, we show that outdoor temperature (varying over time) positively impacts the prediction performance of thermal comfort models by up to 30%. The importance of spatio-temporal context is demonstrated by contrasting micro-level (location specific) and macro-level (6 locations across a city) performance. The most important finding of this work is that a definitive improvement in prediction accuracy is shown with an increase in the time of day and sky illuminance, for multiple thermal comfort metrics

Keywords: Computational Intelligence, Evolutionary Computation, Heuristic Algorithms
Abstract: Abstract—Designing a disassembly line layout is an important part of optimizing an end-of-life product recycling process. Parallel disassembly lines have the characteristics of high disassembly efficiency and can disassemble multiple different products simultaneously. This work formulates a mathematical model for optimizing such lines in terms of disassembly profit and the number of skills. It also proposes an improved bat algorithm based on the Pareto principle to solve the model. In order to verify the effectiveness and feasibility of the proposed algorithm, it is compared with the non-dominated sorting genetic algo- rithm and a decomposition-based multi-objective evolutionary algorithm. Experimental results indicate that this algorithm has outstanding solution capability and is thus suitable for solving parallel disassembly line balance problems.

Keywords: Application of Artificial Intelligence
Abstract: Multi-Agent Path Finding (MAPF) is an important problem with a variety of applications. Its aim is to find collision-free paths for agents having separate start and goal positions. This work proposes a new lexicographic bi-objective MAPF considering different task types, where agents are divided into two kinds to perform critical and acritical tasks. This is common in practical intelligent warehousing scenarios where a critical/acritical-task-performing agent (called c-agent and a-agent, respectively) may represent a full-load/no-load one or the one conducting urgent/non-urgent tasks. The primary objective is to minimize the sum-of-costs of c-agents, while the secondary objective is to minimize the sum-of-costs of a-agents. Two MAPF algorithms are modified to fit and solve the concerned problem for the first time. Moreover, four improvement strategies are embedded to the proposed algorithms and proved to be effective in solving MAPF problems with different task types.

Keywords: Application of Artificial Intelligence, Artificial Life, Biometric Systems and Bioinformatics
Abstract: Early depression detection research employs machine learning models trained on crowd-sourcing data. The training data easily suffer from label noise due to weak self- perception of people and uncontrollability of collection process. The noise issue is seldom discussed in the previous depression detection work. In this work, we firstly introduce the influence based relabeling method in the depression detection task to revise noise labels, and further move one step forward to propose a threshold ratio function to control the relabeling sample size. The relabeling sample size is usually ignored in the previous influence function, so that the relabeling is sometimes overwhelming, leading to great change on the distribution of the training data, and model performance decline. Our proposed method aims at avoiding giant change on the training data. To achieve this, we design an adjustable ratio threshold for the samples to be relabeled. The ratio is adjusted according to the trained model performance. If the model has good performance on the validation set, the relabeling ratio tends mild, otherwise, the relabeling can be aggressive. In the experiments, we recruited 205 participants and collect the usage data from smartphones and wearable bands, including participants’ response to the questionnaire Depression Anxiety Stress Scale- 21. We discuss several main stream denoising methods and compare four most recent methods in the depression detection task. The proposed model achieves a best testing F1 score of 86.3%.

Keywords: Biometric Systems and Bioinformatics, Cloud, IoT, and Robotics Integration
Abstract: Assisting patients with dementia is a significant social issue. Currently, to assist patients with dementia, a multi-modal care technique called Humanitude is gaining popularity. In Humanitude, the patients are assisted through various techniques to stand up independently by utilizing their motor functions as much as possible. Humanitude care techniques encourage caregivers to increase the area of contact with patients during the sit-to-stand motion. However, Humanitude care techniques are not accurately performed by novice caregivers. Therefore, in this study, a smock-type wearable sensor was developed to measure the proximity between caregivers and care recipients during sit-to-stand motion assistance. A measurement experiment was conducted to evaluate the proximity differences between Humanitude care and simulated novice care. In addition, the effects of different care techniques on the center of mass (CoM) trajectory and muscle activity of the care recipients were investigated. The results showed that the caregivers tend to bring their top and middle trunk closer in Humanitude care compared with novice simulated care. Furthermore, it was observed that the CoM trajectory and muscle activity under Humanitude care were similar to those observed when the care recipient stands up independently. These results validate the effectiveness of Humanitude care and provide useful information for teaching techniques in Humanitude.

Keywords: Machine Learning, Image Processing and Pattern Recognition, Neural Networks and their Applications
Abstract: In this paper we propose a cellular diagnostic support system for cervical cancer from WSI which presents the reason for diagnosis. First, cytoplasmic and nuclear regions are extracted from cell images using U-2-Net. For the extracted regions, 29 image features were calculated, and machine learning is applied to perform cell classification based on the Bethesda system. The results showed an accuracy of 68.5% in the 6-class classification of cervical cancer.

Keywords: Application of Artificial Intelligence, Deep Learning, Transfer Learning
Abstract: The number of patients of malignant soft tissue tumor is about 3,000 every year in Japan. In the treatment procedure of soft tissue tumor, survivability of patient must be predicted based on experience. The miss-prediction causes the unnecessary treatment or patient death, an objective decision making system based on pathological image is required. This paper proposes a convolutional neural network (CNN) based survivability and survival time prediction method using pathological image of soft tissue tumor. The proposed method trained Inception v3 and ResNet14-based modified CNN model using 47 pathological images of 26 subjects. As the result of 4-fold cross validation test, the image-wise survivability, image-wise non survivability, subject-wise survivability, and subject-wise non survivability were predicted in F-measure of 0.847, 0.743, 0.909 and 0.824, respectively. Additional experiment using non-survival patients showed that the survival time was predicted in 4.57 months error.

Keywords: Biometric Systems and Bioinformatics, Computational Intelligence, Cybernetics for Informatics
Abstract: The effects of individual physiological phenomena play an important role considering the accuracy of artificial pancreas systems. An example of these phenomena is the heart rate which is easy to measure. There is a connection between heart rate and endogenous glucose production that significantly influences the blood glucose level. The proper implementation of heart rate could lead to defining physical activity in T1DM models. The aim of the current study is to examine how the change in endogenous glucose production influences the fitting accuracies in model versions with different complexity. Our extensions include a heart rate dependent endogenous glucose production equation, modeling the effect of physical activity, and a part defining the effect of insulin on endogenous glucose production. The joint effect of the mentioned extensions was also considered.

Keywords: System Modeling and Control, Robotic Systems, Control of Uncertain Systems
Abstract: The cooperation between humans and robots is of great importance e.g. in medical, industrial or service applications. Here, the task to be pursued by the robot often depends on the current goal of the human. In cases where a direct communication of the human's goal is impractical or even impossible, an estimation of the human's goal is necessary. This estimation as well as potential process or measurement noise introduces uncertainty that needs to be taken into consideration during the planning of the robot's actions.

To this end, we propose an automation comprising an Unscented Kalman filter as state estimator, a model based intention recognition algorithm to estimate the human's goal and a model predictive belief space controller based on Belief i-LQG explicitly considering the estimation uncertainty. We report on a simulated scenario featuring a mobile robot platform cooperating with a human. It demonstrates the ability of the proposed automation to actively reduce uncertainty about the system states and the human's goal while successfully pursuing the overall cooperative task.

Keywords: System Modeling and Control
Abstract: This paper considers the use of a novel canting keel system for the roll stabilization of ships subjected to unbalanced loading occurring, for example, in crane operations at zero surge velocities. A simplified 1 degree-of-freedom mathematical model of a monohull marine vessel based on the lumped parameter approach is considered along with the canting keel system. Then, in order to compensate for unknown plant dynamics as well as time-varying load disturbances, a simple sliding-mode controller is proposed. Simulation results are finally given to illustrate the potential of the approach.

Keywords: System Modeling and Control, Robotic Systems, Control of Uncertain Systems
Abstract: Rearranging deformable objects is a long-standing challenge in robotic manipulation for the high dimensionality of configuration space and the complex dynamics of deformable objects. We present a novel framework, Graph-Transporter, for goal-conditioned deformable object rearranging tasks. To tackle the challenge of complex configuration space and dynamics, we represent the configuration space of a deformable object with a graph structure and the graph features are encoded by a graph convolution network. Our framework adopts a FCN-based architecture to output pixel-wise pick-and-place actions from only visual input. Extensive experiments have been conducted to validate the effectiveness of the graph representation of deformable object configuration. The experimental results also demonstrate that our framework is effective and general in handling goal-conditioned deformable object rearranging tasks.

Keywords: System Modeling and Control, Control of Uncertain Systems
Abstract: The data-driven control of positive systems with constrained state/input is studied, in this work. Employing a finite number of open-loop data and the concept of the MPC approach, optimization problems containing data-based conditions are developed. At the same time, an infinite horizon cost function’s upper bound is minimized online; thus, the data-based control policy can be considered as an optimal control method. The gains of the constrained data-based control signal are calculated online to improve the performance. Applying the data-based state feedback controller to the system yields positive and stable state trajectories with appropriate transient responses. It is guaranteed that the data-driven control scheme is capable of handling constraints. A practical example is simulated to assess the proposed strategy.

Keywords: System Modeling and Control, Robotic Systems, Modeling of Autonomous Systems
Abstract: The dynamic identification of a robot manipulator is a canonical problem in robotics and is usually solved by employing the least squares technique, considering the so-called base parameters of the robot. Such parameters, however, differ from those to which we really want to have access, such as the center of mass and the moments of inertia with respect to this center. Furthermore, the solution obtained by this technique can lead to unsuitable parameters from a physical point of view, which requires a later stage of physical feasibility analysis. In this article, we seek to overcome these drawbacks by identifying the parameters of a 7-DoF lightweight robot using only population-based metaheuristics. We make a study between different techniques and compare them with Bayesian optimization, another black-box approach. We also propose a new algorithm called Probabilistic Differential Optimization and show that this algorithm is more efficient than the others considered and can find a physically feasible solution because of its probabilistic exploration ability. We also study the need for more than one excitation trajectory to overcome measurement noise and overfitting, and finally obtain accurate parameters of the Kinova Gen3 arm equipped with a Robotiq 2F-85 gripper.

Keywords: Human-Machine Cooperation and Systems, Human Factors, Virtual and Augmented Reality Systems
Abstract: Reliability is an important concept contributing to building trust in human-machine systems (HMS). Existing studies have reported separate assessment of human and machine reliability. Thus, there is a gap on considering human and machine reliability concurrently in HMS. To fill the gap, this study investigated the feasibility of such concurrent consideration by using a virtual environment (VE) approach to simulate an HMS. In a developed VE, each human participant performed a task of exploring an invisible surface to perceive its shape, followed by his/her response to a recommendation about the shape made by the VE setting (the machine). Related to human reliability, the perception might be disrupted through a mismatch between the actual shape and force feedback delivered to the participant’s hand. Associated with machine reliability, the recommendation could be incorrect to induce a fault in the setting. Thus, the shape of the invisible surface became an instrument to combine human and machine reliability. The outcomes of the study confirmed the feasibility of combining human and machine reliability in the HMS. Moreover, human reliability might be dominant in the HMS to accomplish the task.

Keywords: Human-Machine Interface, Human Factors, Information Visualization
Abstract: In recent years, maritime operations have become more technologically demanding due to the more complex working condition and stricter safety requirements. The need to improve the performance of human-machine cooperation in navigation through a more intelligent system in path planning, while taking into account the human factors, is more and more urgent. In this paper, a model predictive control (MPC) optimization scheme is proposed for collision-free path planning taking into account the ship dynamics and International Regulations for Preventing Collisions at Sea (COLREGS) explicitly. It utilizes three potential fields which are designed based on COLREGS and the experience of navigators. Different tuning parameter sets are tested in a single encounter scenario including give-way, head-on and overtaking, and multiple ship encounter with 5 target ships is evaluated. The simulation shows promising results where the own ship can perform evasive action according to different encounter types following COLREGS Rule 13-15 and achieve the target positions while maintaining a safe distance during the collision avoidance period. The shortest distances between the own ship and target ships in multiple encounter scenario are all larger than 250m, which further proves the effectiveness of the algorithm.

Keywords: Human-centered Learning, Human Performance Modeling, Intelligence Interaction
Abstract: The global receptive field plays a critical role in visual object tracking. In most popular tracking paradigms, we find that the local receptive field introduced by the convolutional neural network prevents the tracker from focusing on the long-range dependency. Although the Vision transformer brings the global receptive field in downstream tasks, its computational burden remains unaffordable. In this paper, we present a simple yet effective Multilayer Perceptron-based Tracking~(MLPT), including the global receptive field. The MLPT contains three Components: Feature Correlation~(FC) module, Global Information Encoder~(GIE) module and Corner Head(CH). Firstly, the FC module is proposed to effectively converge the template and search region features for generating delicate feature. Secondly, the GIE is designed to integrate the channel and spatial-information dealed with channel-encoding the token-encoding, separately. Specially, the same kernel is utilized for token-encoding in all channels so that our model has the global receptive field. Then, the CH is applied to establish a simple flexible way via computing the box corners coordinates for tracking. Finally, the MLPT, to our knowledge, is the first baseline of MLP-based architecture for object tracking. Extensive experiments are conducted on four challenging datasets, including GOT-10K, LaSOT, UAV123, and TrackingNet. The results show that the proposed method achieve state-of-the-art performance.

Keywords: Virtual and Augmented Reality Systems
Abstract: Global localization in a prior map is an important field in virtual and augmented reality systems, but it is always a challenge to conduct point cloud based localization in the large-scale scene prior map. A large-scale prior map usually means huge amount of calculation for point cloud processing, which leads to the long time required for global localization. To deal with this problem, we propose a fast point cloud global localization method based on point clouds segmentation and learning-based descriptor. On the one hand, cylindrical filtering, ground-point removal and point cloud segmentation are adopted to eliminate a large number of useless points and retain points with rich structures, which improves the efficiency of point cloud registration. On the other hand, reliable 3D point cloud descriptor, two-phase search strategy for place recognition and geometric consistency verification are used to ensure the localization accuracy. Experiments prove that the proposed method achieves good localization effect on both KITTI and MVSEC datasets. Under the condition of ensuring the high localization accuracy, the time for point clouds to complete the global localization is greatly reduced.

Keywords: Human Factors, Human Performance Modeling, Human-centered Learning
Abstract: The development of maritime autonomous surface ships (MASS) has triggered interest from both academia and industry in recent years. Nevertheless, there are several critical phases for fully autonomous ships (MASS-IV); human operators will continue to perform dominant roles onboard for the next decades. The authors conceive a cyber-physical human framework for the experiment design and implementation in the maritime domain: based on multiple experimental platforms with data exchange ports, apply monitoring on and learning from navigators' behaviors, and adapt the ship-bridge system to provide decision support in terms of guidance, navigation, and control. The platforms include a compact simulator from Kongsberg and an immersive simulator for preliminary research design, a group of standard maritime training simulators, and a research vessel. These platforms are utilized for data collection, scenario design, testbeds to demonstrate and verify new techniques and algorithms, etc. The authors illustrate how the framework aids the MASS research and benefits the development process.

Keywords: Brain-based Information Communications, Human-Machine Cooperation and Systems, Assistive Technology
Abstract: Digital platforms could facilitate meditation practice by discriminating the participants’ mental state in real time based on neural activities. However, the search for neural correlates of meditative states yields contradictory results in the literature. To identify the neural signature of meditation, we propose a Random Forest classifier to discriminate between a Focused Attention Meditation (FAM) and a problem-solving task, based on two-second samples of EEG data. Two types of classifiers are considered: individual classifiers, trained on EEG data from the considered subject, and general classifiers, trained on inter-individual data. Our results show that the individual classifiers achieve superior performance with an average accuracy of 93% over 14 subjects. The general classifiers display a lower accuracy (74% and 54% depending on whether the data from the tested subject was included in the training set). This study suggests that automatic detection of meditative processes greatly benefits from intra-personal training. The most discriminating EEG features between the two tasks are the Beta mean band amplitude and the Theta-Gamma phase-amplitude coupling, particularly in the occipital and left centro-temporal brain regions. Our findings favor personalized classifiers for FAM.

Keywords: Human Performance Modeling, Assistive Technology, Human-Machine Interface
Abstract: Due to an aging population and a greater prevalence of diseases like dementia and stroke, there is a growing need for non-invasive and easy to use cognitive training and rehabilitation. Brain-computer interfaces are emerging as a means for such non-invasive cognitive training. This paper proposes a phenomenological model of performance in a P300-speller task, which takes task difficulty into account. The model can be used for simulation purposes and to inform how to adapt the task difficulty in cognitive training or rehabilitation. Inspired by the phenomena of slacking and neural plasticity, a nonlinear autoregressive model with exogenous inputs was trained on, and validated against, the Akimpech dataset, which contains EEG data from healthy subjects who completed several runs of a P300-speller task. The model can simulate or predict the performance in each run of the task. An average R^2 score of 94.15% and 94.11% was achieved on validation data for simulation and 1-step ahead prediction, respectively. The ability of the model to generalize to other experimental setups will be evaluated in the future.

Keywords: Passive BMIs, BMI Emerging Applications
Abstract: Brain Computer Interface (BCI) technology offers the possibility to monitor users’ attention and engagement during learning tasks, enabling adaptation of pedagogical strategies for a personalized learning experience. In this paper, we present an EEG-based passive BCI system for real-time evaluation of user engagement during a language learning task. The EEG Engagement Index, which has been previously associated with attention and vigilance, is measured from three frontal electrodes and used in this system as a neural indicator of engagement. To validate our system, we used it in a human-robot interaction (HRI) setting, in which a robot tutor monitored the learner’s brain activity and adapted its tutoring strategy when a lapse in engagement was detected. We discuss the challenges and preliminary results from our pilot study with eight participants.

Index Terms—Brain computer interface (BCI), Adaptive learning, EEG Engagement Index, Educational agents, Human- robot interaction (HRI), Robot-assisted Language Learning (RALL)

Keywords: Other Neurotechnology and Brain-Related Topics, BMI Emerging Applications, Passive BMIs
Abstract: Meditation methods, which have their origins in ancient traditions are gaining popularity as a result of their potential mental and physical health advantages. EEG neu- ral correlates underlying enhanced cognitive abilities such as sustained attention and working memory need to be analyzed scrutinizingly to evaluate the effects of meditation practices. In this article, we thus provide an analysis of EEG features such as various band powers and connectivity based features to evaluate the meditation effects. Also, we provide a classification framework to classify the meditation states from the baseline EEG states. We report our results on an in house dataset of 20 participants(10 experienced and 10 novice) who underwent a two-week long mantra meditation practice. Strikingly we have found out that, as the novice participants practice meditation overtime, the accuracies of machine learning classification between the baseline EEG versus meditative EEG of the novice increase significantly, as an indication of their enhanced meditation experiences. Also, we found out that even such short and regular meditation practices, the cognitive abilities of novice meditators get enhanced which are evaluated through Brain-Based Intelligence Test (BBIT) psychometric tests and the results that got reflected in their EEG correlates.

Keywords: Other Neurotechnology and Brain-Related Topics, BMI Emerging Applications, Passive BMIs
Abstract: Human-machine systems (HMS) need trustful cooperation between humans and machines for achieving a goal. Establishing such trust demands the machines’ adaptivity to the cognitive workload (CWL) of the humans. The CWL is conventionally measured as self-reported scores from a NASA-TLX questionnaire, susceptible to individual subjectivity. In contrast, logged brainwaves are useful for measuring the CWL objectively. However, there is a literature gap of mapping the brainwaves to a CWL factor – i.e., physical demand. As a feasibility, we thus proposed an activity ratio (AR) to measure the physical demand from the brainwaves. Statistical analyses indicated significant correlations between the AR and self-reported scores of the physical demand, compared to a well-known engagement ratio. This finding implied the feasibility of the AR to measure the physical demand.

Keywords: Intelligent Power Grid, System Modeling and Control, Electric Vehicles and Electric Vehicle Supply Equipment
Abstract: With the development of renewable energy, much attention has been paid to improving energy efficiency. This paper proposes a residential microgrid scheduling method to improve the energy utilization rate of residential buildings. Firstly, the power cost model for residential users and the specific charge-discharge models of electric vehicles and energy storage equipment is constructed. Then the whole energy scheduling process is formulated as a multi-variable mixed integer linear programming (MV-MILP) problem, whose objective is to minimize the power cost of end-users. The differential evolution algorithm is adopted to solve the problem, and the scaling factor adaptation is further used to accelerate the convergence. Simulation results demonstrate the effectiveness of the proposed scheduling method.

Keywords: Smart Buildings, Smart Cities and Infrastructures, Consumer and Industrial Applications, Smart Metering
Abstract: The consumption disaggregation is of paramount importance to the management and efficient use of electricity. Thus, new supervised and unsupervised approaches to nonintrusive load monitoring have been proposed by research groups around the world. In this sense, the present paper proposes an approach based on the DAMICORE clustering algorithm. For that, an experimental bench composed of six appliances was used, which had their individual and combined current signatures acquired. Therefore, scenarios in which there is an overlap in the consumption were considered. Based on the current measurements, the signals were windowed (1 and 1/2 cycle) and, subsequently, features at the time and frequency domains were extracted. These features were then submitted to the clustering stage to identify the state of each appliance. Due to the balancing of the dataset, the accuracy was calculated to determine the DAMICORE performance. Thus, it was possible to observe that the use of 1-cycle windows allowed obtaining a better average accuracy (80.16%) when using all extracted features. However, when considering only the most relevant feature types per appliance, there was an improvement in the average accuracy, reaching 87.83% and proving to be better than the state-of-the-art approaches.

Keywords: Smart Buildings, Smart Cities and Infrastructures, Decision Support Systems, Infrastructure Systems and Services
Abstract: Accurate and real-time prediction of water quality not only helps to assess the environmental quality of water, but also effectively prevents and controls water quality emergencies. In recent years, neural networks represented by Bidirectional Long Short-Term Memory (BiLSTM) and Encoder-Decoder (ED) frameworks have been shown to be suitable for the prediction of time series data. However, traditional statistical methods cannot capture nonlinear characteristics of the water quality, and deep learning models often suffer from gradient disappearance and gradient explosion problems. This work proposes a hybrid water quality prediction method called VBEG, which combines Variational Mode Decomposition (VMD), BiLSTM, an ED structure, and Genetic Simulated annealing-based particle swarm optimization (GSPSO). VBEG first adopts VMD to deal with nonlinear features in the original time series. Then, VBEG combines BiLSTM and the ED structure to capture bi-directional long-term correlations, and realize dimensionality reduction, respectively. Furthermore, VBEG adopts GSPSO to optimize its hyperparameters. Experimental results with real-life datasets demonstrate that the proposed VBEG outperforms two current state-of-the-art algorithms in terms of prediction accuracy.

Keywords: Smart Buildings, Smart Cities and Infrastructures, Infrastructure Systems and Services, Decision Support Systems
Abstract: The water quality data has missing values and lacks integrity because water environment monitoring equipments are easily damaged by environmental influences, thereby affecting the analysis accuracy of downstream tasks. Traditional data imputation methods include mean/last filling, K-nearest neighbor, matrix factorization, Lahrangian interpolation, etc., do not consider time dependence or fail to use complex relations among multiple features. Inspired by successful applications of various variants of Generative Adversarial Networks (GANs) on time series data, this work proposes a time series data imputation method called GEDA, which integrates GAN, an Encoder-Decoder structure, and an Autoregressive network. GEDA adopts GAN to learn the probability distribution of multi-feature time series, and imputes the missing values with the generated data. Then, GEDA combines feature extraction and dimensionality reduction of the encoder-decoder structure, and time dependence capturing of the autoregressive network. Real-life dataset-based experimental results demonstrate GEDA outperforms several state-of-the-art data imputation methods in terms of accuracy.

Keywords: Infrastructure Systems and Services, Smart Buildings, Smart Cities and Infrastructures, Discrete Event Systems
Abstract: Accurate prediction of water quality indicators can effectively prevent sudden water pollution events, and control pollution diffusion. Neural networks, e.g., long short-term memory (LSTM) and encoder-decoder network, have been widely used to predict time series data. However, as the water quality data increases, it becomes unstable and highly nonlinear. Accurate prediction of water quality becomes a big challenge. This work proposes a hybrid prediction method called VBAED to predict the water quality time series. VBAED combines Variational mode decomposition (VMD), Bidirectional input Attention mechanism, an Encoder with bidirectional LSTM (BiLSTM), and a Decoder with temporal attention mechanism and LSTM. Specifically, VBAED first adopts VMD to decompose the ground truth time series, and the decomposed results are used as the input along with other features. Then, a bidirectional input attention mechanism is adopted to add weights to input features from both directions. VBAED adopts BiLSTM as an encoder to extract hidden features from input features. Finally, the predicted result is obtained by an LSTM decoder with a temporal attention mechanism. Real-life data-based experiments demonstrate that VBAED obtains the best prediction results compared with other widely used methods.

Keywords: Infrastructure Systems and Services
Abstract: Content

- Overview by Christopher Gorog, Initiative Chair
- Aspects of Privacy by Vladik Kreinovich, IEEE SMCS Representative at the Initiative
- General Discussion and Q&A by Wlodzislaw Duch, Chris Gorog, and Stefano Zanero

Keywords: Technology Assessment
Abstract: Working lunch of Associated Editors together with IEEE SMCS VP Publication and IEEE SMCS EICs

Keywords: Technology Assessment
Abstract: Presentation of the five best student paper candidates (selected by the IEEE SMC Best Conference Paper Awards Subcommittee)

• Candidate 1 (Tu-PS1-T8.1): State Observer for Position Control of Systems with Quantized Outputs in Large Scale Robotics" (Müller, Bernd)
• Candidate 2 (Mo-PS2-T5.1): "Ant Colony Optimization for Electric Vehicle Routing Problem with Capacity and Charging Time Constraints" (Nie, Zihao)
• Candidate 3 (Mo-PS3-T10.4): "Segmentation of Indoor Daily Living Environments into Regions Used for Different Purposes" (Naito, Kenji)
• Candidate 4 (Tu-PS2-T10.3): "A Passive Brain-Computer Interface for Monitoring Engagement During Robot-Assisted Language Learning" (Prinsen, Jos)
• Candidate 5 (Mo-PS4-T2.5): "DandelionTouch: High Fidelity Haptic Rendering of Soft Objects in VR by a Swarm of Drones" (Fedoseev, Aleksey)

Keywords: Human-Machine Cooperation and Systems, Human Factors, Human Performance Modeling
Abstract: Digital human models (DHM) offer a great possibility of design evaluation for products humans interact with. One field to apply DMH's is the dynamic analysis of human-machine interaction. Thereby, the dynamics of the human body can significantly influence the entire system behavior and thus it is important to consider during the design process. For the modeling language Modelica no simulation library for this application is available, so far. The language is especially suitable for the simulation across different domains which makes a DHM in Modelica useful. In this paper a DHM predicting the dynamic behavior of the human body during human-machine interaction is implemented in Modelica. It consists of a multi-body skeleton model, an inverse kinematics algorithm for the limbs and a controller for the skeleton joints inspired by human motor control. The capability of the proposed simulation approach to take the relevant dynamical properties of the human body and motor control into account is demonstrated in a use-case simulation. As use-case the aircraft-pilot coupling during a strong gust acting on a sports aircraft is investigated.

Keywords: Human Factors, Human-Computer Interaction, Human-Machine Cooperation and Systems
Abstract: During the pandemic, people gradually realized the limitations of remote problem-solving collaboration. Thus far, the impact of online platforms on problem-solving cognitive processes has not been thoroughly investigated, despite the active use of such processes in remote work. This study analyzes the differences in cognitive patterns between online and offline problem-solving meetings. Discussion data containing approximately 5,000 utterances were subjected to entropy and content analysis. The results showed a distinct difference in cognitive patterns in discussions conducted on various platforms, as online discussions require more cognitive effort for technological equipment operation. Online meeting solutions were found to be less developed than those generated in offline meetings. According to the results obtained, we propose a series of high-level platform-neutral steps for small-scale problem-solving. The findings not only contribute to building platform-specific facilitation guidelines, but they also aid research on human-computer interaction from the perspective of online discussion platform designs and methods for cognitive process evaluation.

Keywords: Human-Machine Cooperation and Systems, Human Factors, Human Performance Modeling
Abstract: Building on the benefits of Industry 4.0, Industry 5.0 promotes human-centricity of factories, placing operator's wellbeing at the center of production. Production processes are expected to be designed around operator's needs, for a more sustainable contribution of industry to society. In this broad context, in this paper we consider the problem of monitoring operator's condition and, specifically, detecting any mental or physical fatigue they might be experiencing in workplaces. Indeed, if the onset of any source of fatigue is monitored, it is possible to introduce assistive strategies that preserve productivity, on the one side, and operator's wellbeing, on the other side. To achieve this goal, we detect mental and physical fatigue conditions via physiological monitoring, by means of a wearable device that measures cardiac activity. We design an experimental protocol such that test participants are exposed to mental and physical fatigue. Their heart rate variability is then extracted and analysed to discriminate among rest, mental fatigue, physical fatigue and joint mental and physical fatigue. The achieved results show that statistically significant differences can be found in time-domain metrics. Moreover, the analysis of the empirical distribution functions shows, for each metrics, the conditions that exhibit the greatest differences and, hence, that can be distinguished more accurately. However, results show also that, in the presence of physical fatigue, it is difficult to detect the presence of additional mental fatigue.

Keywords: Human-Machine Cooperation and Systems, Human-Machine Interface, Virtual and Augmented Reality Systems
Abstract: Quadruped platforms have become an active topic of research due to their high mobility and passability in rough terrain. However, it is highly challenging to determine whether the clattered environment could be passed by the robot and how exactly its path should be calculated. Moreover, the calculated path may pass through areas with dynamic objects or environments that are dangerous for the robot or people around. Therefore, we propose a novel conceptual approach of teaching quadruped robots navigation through user-guided path planning in virtual reality (VR). Our system contains both global and local path planners, allowing robot to generate path through iterations of learning. The VR interface allows user to interact with environment and assist quadruped robot in challenging scenarios. The results of comparison experiments show that cooperation between human and path planning algorithms can increase the computational speed of the algorithm by 35.58% in average, and non-critically increasing of the path length (average of 6.66%) in test scenario. Additionally, users described VR interface as not requiring physical demand (2.3 out of 10) and highly evaluated their performance (7.1 out of 10). The ability to find a less optimal but safer path remains in demand for the task of navigating in a cluttered and unstructured environment.

Keywords: Human-Computer Interaction, Human-Machine Interface, Intelligence Interaction
Abstract: Nowadays, autonomous mobile robots support people in many areas where human presence either redundant or too dangerous. They have successfully proven themselves in expeditions, gas industry, mines, warehouses, etc. However, even legged robots may stuck in rough terrain conditions requiring human cognitive abilities to navigate the system. While gamepads and keyboards are convenient for wheeled robot control, the quadruped robot in 3D space can move along all linear coordinates and Euler angles, requiring at least 12 buttons for independent control of their DoF. Therefore, more convenient interfaces of control are required.

In this paper we present HyperPalm: a novel gesture interface for intuitive human-robot interaction with quadruped robots. Without additional devices, the operator has full position and orientation control of the quadruped robot in 3D space through hand gesture recognition with only 5 gestures and 6 DoF hand motion.

The experimental results revealed to classify 5 static gestures with high accuracy (96.5%), accurately predict the position of the 6D position of the hand in three-dimensional space. The absolute linear deviation Root mean square deviation (RMSD) of the proposed approach is 11.7 mm, which is almost 50% lower than for the second tested approach, the absolute angular deviation RMSD of the proposed approach is 2.6 degrees, which is almost 27% lower than for the second tested approach. Moreover, the user study was conducted to explore user's subjective experience from human-robot interaction through the proposed gesture interface. The participants evaluated their interaction with HyperPalm as intuitive (2.0), not causing frustration (2.63), and requiring low physical demand (2.0).

Keywords: Consumer and Industrial Applications, Communications, Electric Vehicles and Electric Vehicle Supply Equipment
Abstract: The new manufacturing paradigm, which is organized under Industry 4.0, has strict requirements for supporting technologies. Wireless communication technology should be available to support movable nodes and to make a production site flexible and able to adapt to the various requirements of the manufacturing operations and the production management systems. Depending on the application and processes being serviced, communication is subject to a broad range of strict requirements, from reliability and predictable delays to low cost and reduced energy consumption. In this paper, we present a real scenario in which Autonomous Guided Vehicles (AGVs) equipped with Collaborative robots (Cobots) are used as transportation and production units, and therefore, a set of communication services must be implemented. The challenges associated with moving AGVs are discussed, and the associated requirements that apply to wireless communication are presented. The results allow to formulate a definition of a three-level hierarchy for communication services (onboard, shopfloor, and system) that are required to implement the Industry 4.0 paradigm. In addition, a review of the relevant wireless network protocols that can be used for the implementation of communication services is presented.

Keywords: Intelligent Transportation Systems, Electric Vehicles and Electric Vehicle Supply Equipment, Intelligent Power Grid
Abstract: 近年来，电动汽车（EV）在环境污染方面受到了极大的关注。在这里，与传统的内燃机（ICV）相比，电动汽车可以大大减少环境污染。然而，由于电动汽车不能像ICV那样快速补充，充电基础设施的刚性部署及其有限的充电能力导致服务拥堵，特别是由于大量电动汽车停放充电需求。与在位置和充电便利性方面具有刚性延伸的CS相比，车对车（V2V）充电服务在灵活性方面提供了空间和时间上的进步，有可能补充G2V充电模式，可以补充甚至取代G2V充电模式。本文提出了一种混合V2V充电方案，由直接和异步V2V充电模式组成，以实现极大的充电灵活性，减轻电网负载负担。在这里，我们估计每种模式下的最短等待时间（MWT），作为在模式之间切换和优化ે

Keywords: Robotic Systems, Infrastructure Systems and Services
Abstract: Coverage path planning (CPP) is useful in tasks like map construction and criminal capture. As an outstanding method, predator-prey coverage path planning (PPCPP) employs a predator-prey mechanism to enable a robot to adaptively cover an arbitrary 2-D surface with dynamic obastacles. However, it is designed for bounded environments only and cannot work in unbounded environments. Inspired by the foraging behavior of herds in nature, this work proposes an adaptive coverage path planning algorithm suitable for unbounded environments, called herd foraging-based coverage path planning (HFCPP). Based on the attraction between an animal and its herd, HFCPP employs a virtual herd to control the overall coverage direction of a robot, allowing it to be applied in unbounded environments. The experimental results demonstrate the effectiveness of HFCPP in unbounded environments.

Keywords: Robotic Systems, Cooperative Systems and Control, Mechatronics
Abstract: Advances in robotics and multisensory displays allow extending telepresence ambitions beyond only “the feeling of being present at a remote location”. In this paper, we discuss what may lie beyond telepresence and how we can transport both the functional and social self of a user. We introduce the embodiment illusion and its potential contribution to task performance and list important cues to evoke this illusion, including synchronicity in multisensory information, a first-person visual perspective, and a human-like visual appearance and anatomy of the telepresence robot. We also introduce the concept of social presence and the important bidirectional social cues it needs, including eye contact, facial expression, posture, gestures, and social touch. For all these multisensory and social cues, we explain how they can be implemented in a telepresence system and describe our solution consisting of a closed control pod and a humanoid telepresence robot.

Keywords: Intelligent Transportation Systems, Cooperative Systems and Control, Consumer and Industrial Applications
Abstract: Automated guided vehicles (AGV) provide a cost-efficient transportation method in smart industrial plants. Their continuous operation is crucial for production flow. However, while detection of typical failures, e.g., those related to battery voltage, can be performed in an automated manner, more complex scenarios require expert knowledge and human monitoring. In this paper, we evaluate recurrent neural network-based (RNN) energy consumption forecasting using other telemetry features. We aim to find models well suited for anomaly detection methods working on the analysis of error between forecasted and actual values. We compare the results of RNN architectures on our data and public vehicle energy datasets. We demonstrate that RNN-based forecasting, together with a proper selection of telemetry features used in prediction, can be effectively utilized on AGV telemetry data as a first step in anomaly detection schemes.

Keywords: Optimization and Self-Organization Approaches, Computational Intelligence
Abstract: Trusted requirements are the requirements that affect the trusted attributes of software, which have an important impact on whether trusted software can meet the trusted requirements. However, despite the continuous development of society and technology, the work of obtaining and analyzing trusted requirements has not become easy, has become more and more difficult. In the process of obtaining trusted requirements, a series of problems, such as low efficiency and inaccurate acquisition of requirements, serious software quality problems, budget overruns and delivery delays, have become increasingly prominent. In view of the above problems, this paper integrates the concept of meta-requirement into trusted requirement. Based on the respective characteristics of meta-requirement and trusted requirement, this paper puts forward the concept of trusted meta-requirement, and introduces the basic elements and characteristics. Then the trusted meta-requirements and some rules involved in it are formalized by the combination of first-order logic and set theory in order to improve the accuracy of the description and analysis of trusted requirements.

Keywords: Information Assurance and Intelligence, Cloud, IoT, and Robotics Integration
Abstract: Remote Attestation (RA) is a security service by which a Verifier (Vrf) can verify the platform state of a remote Prover (Prv). However, in most existing RA schemes, the Prv might be vulnerable to denial of service (DoS) attacks due to the interactive challenge-response methodology while there is no authentication about the challenge. Worse, many schemes cannot effectively detect mobile malware that can be inactive during the on-demand attestation launched by the Vrf. In this paper, we propose a self-measurement RA for SGX-based platforms, which can effectively mitigate DoS attacks and defend against mobile malware. To this end, a two-way identity authentication is first enforced between the Prv and Vrf with the help of a blockchain system, in which a shared session key is also generated. Secondly, trigger conditions of measurements on the Prv’s side are time points generated by the Prv self instead of Vrf’s requests. The Vrf can retrieve multiple self-measurement results during one execution of the protocol to monitor the Prv’s platform over a period of time continuously, which can detect mobile malware effectively. Our scheme utilizes SGX to provide the runtime protection for sensitive information such as session key, self-measurement code, time points of self-measurements, and self-measurement results, making a higher security guarantee. In addition, the session key, time points of self-measurements, and self-measurement code can be changed or upgraded, making our scheme more flexible and scalable. The simulation implementation and results show that our scheme is feasible and practical.

Keywords: Neural Networks and their Applications, Deep Learning, Transfer Learning
Abstract: Insulator fault diagnosis is a daily but key task for the power transmission system. Long-term exposure to complex natural environment will cause different insulator defects. As a common defects, missing-cap defects of insulators will not only affect the structural strength of power insulators, but also bring a certain effect to the stable power transmission. With the rapid development of machine learning, some machine learning-based defect recognition methods have been proposed for fast and high-precision power inspection. However, the handcrafted features could not effectively express the aerial images against complex inspection environment to affect detection performance of the shallow learning algorithms. And the detection precision of deep learning algorithms will be affected by the unbalanced small-scale defects. Therefore, the fast and high-precision power inspection still faces a certain challenge in the smart grid. To address the above issues, fusion with the deep convolutional neural network (DCNN) and transfer learning, a novel fault diagnosis algorithm of power insulators is proposed to provide a fast and accurate power inspection scheme. To remove complex backgrounds, a fast insulator location algorithm based on the lightweight YOLOV4 model is proposed which is served for the following defect recognition. On the basis, to imitate human vision, a defect recognition algorithm is proposed based on multi-feature fusion. Meanwhile, to ensure the feature expression ability of transfer learning on power insulators, a novel optimization strategy of transfer learning is proposed to improve the recognition precision. Experiments show that the proposed method could acquire a good recognition performance than other recognition models.

Keywords: Image Processing and Pattern Recognition, Machine Learning, Neural Networks and their Applications
Abstract: Bolts are important parts of transmission lines, and their states are closely related to the safe operation of transmission lines. Compared with insulators, U-shaped rings, and other fittings, bolts are small objects, and the detection of bolt and pin missing is difficult for transmission line patrol video analysis. This paper optimizes the parameters of the YOLOX network to adapt to the multiscale object detection task in the complex scene of the transmission line, and proposes a cascaded network model. The cascaded network locates the important large-scale fittings with bolts by the first layer and then detects the small-scale bolts on the fittings by the second layer, which greatly improves the detection accuracy of the small bolts. Finally, the effectiveness of the method is verified by experiments. The experimental results show that the cascaded YOLOX network can accurately detect the bolts that are originally difficult to detect. It effectively solves the problems of low detection rate of small bolts and pin missing defects.

Keywords: Application of Artificial Intelligence, Deep Learning, Neural Networks and their Applications
Abstract: The matching methods of question pair in the most FAQ system are less optimized for the specific domains, where proper nouns and irregular expressions always exist in the questions. To address the above problems, we propose a knowledge filtering method and the FK-BERT (FAQ-oriented knowledge-enabled BERT) model, which make full use of the domain knowledge graph. In the knowledge filtering stage, the semantic relationships between candidate entities and question sentences are thoroughly evaluated to choose the applicable entities. Furthermore, FK-BERT model considers both the relationships between entities within a single question, as well as the relationships of entities between two questions. The experimental results show that the joint use of knowledge filtering and FK-BERT model can improve the performance of FAQ systems for the domain of operating system.

Keywords: Optimization and Self-Organization Approaches, Neural Networks and their Applications, Machine Learning
Abstract: Noise plays an important role in the gradient-based optimization methods, and a series of numerical experiments have demonstrated that adding gradient noise improves learning for neural networks. However, the mathematical interpretation of the noise remains a challenge. In this paper, we show that, the noise variation can be regarded as a smoothing factor, and we prove that, under certain conditions, a noisy gradient descent (NG) enjoys linear global convergence in expectation sense. We contribute to this problem by introducing an intermediate which connect the NG method to the smoothed function. On the one hand, this connection reveals that applying the NG method to a function is the same as applying the gradient method to the corresponding function smoothed by the noise; and on the other hand, it allows us to establish the convergence behavior of the NG in a global sense. Moreover, we also consider what conditions make the global minimizer of the smoothed function not far from the original global minimizer.

Keywords: Quantum Cybernetics, Machine Learning, Computational Intelligence
Abstract: Precise and resilient quantum gate design is important for the building of quantum devices. In this paper, we consider the optimal and robust quantum gate design problem for three classes of two-level quantum systems. The aim is to construct quantum gates in a given fixed time with limited control resources. A modified dueling deep Q-learning (MDuDQL) is employed for the optimal and robust gate design problem. To improve the performance of the classical DuDQL method, we propose a unique semi-Markov DuDQL algorithm based on a modified action selection procedure, modified replay memory, and soft update procedure. The proposed algorithm outperforms ordinary DuDQL in terms of discovering optimal global or near global optimal control protocols and faster convergence to a better policy. Moreover, the modified DuDQL agent shows improved performance in finding robust control protocols which achieve high-fidelity quantum gate design for varying uncertainties in a certain range. The effectiveness of the proposed algorithm for the optimal and robust gate design problems has been illustrated by numerical results.

Keywords: Cybernetics for Informatics, Machine Learning, Computational Intelligence
Abstract: This paper investigates the application of an evolutionary algorithm, evolution strategy (ES)(μ+λ) to the control design in several inhomogeneous open quantum ensembles. We apply the ES(μ + λ) to assist the sampling-based learning control (SLC) technique, by which a set of control signals is designed to drive the inhomogeneous open quantum ensemble to a given target state. We illustrate our algorithm in two-level and four-level inhomogeneous open quantum ensembles. Numerical results show the effectiveness of the proposed control algorithm. The comparison with other evolutionary algorithms such as differential evolution (DE) and genetic algorithm (GA) shows the superiority of our ES(μ + λ) both in average fidelity and stability. In a four-level open quantum ensemble, for example, the fitness error after optimization using the ES(μ + λ) is decreased by around 59% compared to DE, and the standard deviation is lowered by about 47%.

Keywords: Quantum Cybernetics, Application of Artificial Intelligence, Evolutionary Computation
Abstract: Portfolio optimization is a paramount and important issue in the financial technology area. Constructing a portfolio requires simultaneous considerations of both return and risk, and the proposed system utilizes the trend ratio based on funds standardization, which can effectively evaluate the performance of a portfolio by its returns and risks, while the quantum-inspired tabu search (QTS) algorithm can be used to efficiently build the best portfolio with a stable uptrend. As it is difficult for users to observe a large amount of stock data, this system helps investors to deal with portfolio optimization issues and further provides an information visualization interface to analyze more investment situations. The user interface of the proposed system provides a new service for users to get started quickly in stock selection and clearly analyze the performance of portfolios in the different stock markets. The proposed method can also push notifications automatically when the system detects that the trend of the user’s portfolio is going to slow down, and reminds them to invest in new portfolios with a strong uptrend in a timely manner. The system also designs three games. The first two games make people familiar with stocks. The third game combines social computing with gameplay, which helps to simultaneously find a better portfolio in the highly complex data. In addition, our system provides the functions of general financial applications, such as stock price inquiry, financial news, etc.

Keywords: Quantum Cybernetics, Evolutionary Computation, Computational Intelligence
Abstract: Reversible circuits are the essential building blocks of quantum computers and have zero energy dissipation. However, there is no general rule on how to synthesize an effective circuit with minimal cost. Many researchers have placed a high value on the design of algorithms for reversible circuit synthesis as it is the fundamental component to implement in many paradigms, such as Shor’s algorithm. In this paper, the knowledge navigated quantum-inspired tabu search algorithm (KNQTS) is proposed to synthesize several benchmark functions. KNQTS is a quantum-inspired algorithm with the concept of getting closer to the best solution and keeping away from the worst solution, and it has a great search ability. Furthermore, the proposed algorithm uses self-adaptive, global-best guided and QuantumNot gate mechanisms to make all procedures more efficient and avoid sticking to the local optimum. This paper also compares the KNQTS approach’s experimental results with those obtained via other metaheuristics and previous algorithms. Finally, the result shows that the proposed KNQTS method outperforms other state-of-the-art methods and achieves the same functionality at a lower cost. The solutions are optimal or near-optimal.

Keywords: Application of Artificial Intelligence, Swarm Intelligence, Evolutionary Computation
Abstract: The energy consumption and environmental pollu- tion caused by fuel vehicles have attracted great attention, and now governments around the world are vigorously promoting the production and use of electric vehicles. Among them, the unreasonable layout of charging piles brings inconvenience to users and seriously restricts the development of electric vehicles. In order to deploy charging plies in the existing parking lots more reasonably, this paper intends to propose a fine-grained layout model based on the Voronoi diagram and further develop an adaptive particle swarm optimization (APSO) approach. First, a point-based method is used to estimate charging demands. Second, combined with the Voronoi diagram, this paper proposes a fine-grained layout model of charging piles in parking lots with the goal of maximizing the social benefit. Third, the APSO is developed to solve the optimization problem. Finally, the performance of the model and the proposed algorithm is analyzed through an example in terms of the satisfaction of user needs, the benefits of parking lots, and the density of the charging piles. The numerical experimental results verify the universality and rationality of the proposed model, and thus the proposed model can provide a certain decision-making basis for theoretical and practical research into the new energy vehicle industry.

Keywords: Optimization and Self-Organization Approaches, Swarm Intelligence
Abstract: Latent Factor (LF) models are highly effective in representing high-dimensional and incomplete (HDI) matrices. Hessian free (HF) optimization is an efficiency second-order algorithm to minimize the object function in LF models. An HF-based second-order LF model can achieve better accuracy representation results than first-order ones with affordable computational burden. However, its low rank representation ability relies on a more complex training process, multiple hyper-parameters work cooperatively to decide the results. Thus, these hyper-parameters should be turned with care since they are mutually influenced. The heavy hyper-parameters turning work reduces the practicability of a second-order LF model. To address this issue, this study incorporates the principle of particle swarm optimization into the second-order LF model to propose an adaptive second-order LF (ASLF) model. Experimental results on three HDI matrices reveal that ASLF model can be fine-tuned adaptively with acceptable computation burden.

Keywords: Swarm Intelligence, Computational Intelligence
Abstract: This paper extends boolean particle swarm optimization to a multi-objective setting, to our knowledge for the first time in the literature. Our proposed new boolean algorithm, MBOnvPSO, is notably simplified by the omission of a velocity update rule, and also has enhanced exploration ability due to the inclusion of a "noise" term in the position update rule that prevents particles being trapped in local optima. Our algorithm additionally makes use of an external archive to store non-dominated solutions and implements crowding distance to encourage solution diversity. In benchmark tests, MBOnvPSO produced high quality Pareto fronts, when compared to benchmarked alternatives, for all of the multi-objective test functions considered, with competitive performance in search spaces with up to 600 discrete dimensions.

Keywords: Evolutionary Computation, Swarm Intelligence, Computational Intelligence
Abstract: 由于传统烟花算法（FWA）容易陷入局部优化，全局优化能力不足，优化精度低，因此提出基于粒子群优化和小孔成像逆向学习的增强型烟花算法（EFWA-SI）。首先，设计了基于种群演化速度的动态爆炸振幅机制。通过评估每个迭代过程中适应度的进化速度，动态调整群体的进化速度，并控制全局和局部搜索信息。其次，提出一种结合最优-最差逆向学习思路的小孔成像逆向学习策略。该策略增加了最优求优位置的多样性，提高了算法跳出局部最优值的能力;最后，根据每个烟花的位置和速度更新当前位置，同时更新全局最优烟花和个人最优烟花，提高个人启发式学习能力。通过CEC2015测试函数的仿真实验，并与相关算法进行对比，结果表明，该算

Keywords: Swarm Intelligence, Evolutionary Computation, Computational Intelligence
Abstract: This paper proposes a ranking weight-based roulette wheel selection (RWRWS) method for a promising particle swarm optimizer, called comprehensive learning particle swarm optimizer (CLPSO), to further improve its optimization performance. Specifically, the proposed RWRWS adopts a non-linear weight function to enhance the selection probabilities of promising personal best positions during the exemplar construction. In this way, it is expected that the construction efficiency of generating a promising leading exemplar for each particle could be improved and thus the optimization performance of CLPSO is expectedly elevated. To validate the feasibility and effectiveness of RWRWS, we carry out extensive experiments on a widely acknowledged benchmark problem set by comparing it with other three selection methods, namely the fitness-based roulette wheel selection (FRWS), the ranking-based roulette wheel selection (RRWS), and the tournament selection (TS). Experimental results demonstrate that RWRWS helps CLPSO attain the best overall performance among the four selection methods.

Keywords: Neural Networks and their Applications, Deep Learning
Abstract: The rapid development of social media makes it easy for people to acquire information while also providing a platform for publishing and spreading fake news. Fake news brings plenty of explicit and implicit risks to social stability, making fake news detection an issue that deserves attention. Recent methods based on graph neural networks (GNN) achieve impressive results in fake news detection, but their performance is still limited in practice due to the absence of high-order relations between nodes. In this paper, we propose a SentimentAware Hypergraph Attention Network (SA-HyperGAT) for fake news detection, SA-HyperGAT can better leverage different kinds of information from news contents and user comments with hypergraphs, which can capture higher-order dependency between words and sentences compared with general graphs. Specifically, we first construct two hypergraphs with distinct types of nodes and hyperedges to utilize structural information of news contents and sentimental information of user comments. Then we adopt a hypergraph attention network with a dual attention mechanism to learn the composed representations of two hypergraphs for the final prediction. Our proposed SAHyperGAT outperforms competitive baselines on two real-world datasets. Extensive experimental results prove the effectiveness of each component in SA-HyperGAT.

Keywords: Deep Learning, Computational Intelligence
Abstract: Sarcasm is a sophisticated expression, commonly used on social media, e.g., Reddit and Twitter. The presence of sarcasm in social media text flips the polarity of sentiment, thus hindering the performance of works that require true sentiment, e.g., sentiment analysis and opinion mining. However, current works fail to exploit commonsense knowledge in the sarcasm detection task. In this paper, we revisit sarcasm detection from a novel perspective, which models commonsense knowledge as well as context semantics to reason with sarcasm. More specifically, we propose a commonsense-aware model with a heterogeneous graph attention network that leverages commonsense knowledge, enabling it to better understand implied sentiment behind the literal meaning. We conduct experiments on benchmark datasets from Reddit and Internet Argument Corpus. Experimental results show that our proposed approach yields superior performance with commonsense knowledge integrated.

Keywords: Application of Artificial Intelligence, Machine Vision, Deep Learning
Abstract: Optical flow estimation is an essential task in computer vision. Standard cameras are prone to blurred images or over-saturated regions under extreme conditions. The event camera is a novel vision sensor inspired by the biological retina. It has the advantages of high time resolution, low delay and high dynamic range. We propose a new event representation and a novel deep learning network E-HANet (Event-based Hybrid Attention Network) for event-based dense optical flow estimation. To take full advantage of the complementarity between positive and negative events, we introduce the stacked positive and negative event slices as input. The feature extractor based on the channel attention is able to model the features from different event slices and fuse them by weight. We then present the hybrid attention weighting module to globally aggregate motion features. Compared to the event-based state-of-the-art, our approach reduces the average end-point error by 3% on the MVSEC dataset and 20% on the DSEC-Flow dataset.

Keywords: Machine Vision, Image Processing and Pattern Recognition, Deep Learning
Abstract: As a fine-grained semantic segmentation task, human parsing has attracted extensive attention in computer vision. However, without the assistance of heterogeneous information, it is difficult to obtain detailed human parsing directly. At present, although some studies have introduced heterogeneous data, such as pose estimation and edge prediction, to guide human parsing task, the correlations between these heterogeneous data has not been effectively utilized. To avoid the distribution gap among heterogeneous data, we proposed a Heterogeneous Interactive Attention Network (HIANet), in which we exploit the attention between heterogeneous data to capture long-distance context dependence. And the supplementary cues with plentiful interaction can mutually guide multi-source features to correct their respective prediction errors, further refine the result of human parsing. Extensive experiments on three human body parsing datasets are conducted, especially on the LIP dataset, where the mean accuracy (mean Acc) and mean Intersection-over-Union (mIoU) of the proposed HIANet are improved by 2.86% and 3.90% compared with PGECNet,respectively. Our code has been made available at https://github.com/wangwenjiawj/HIANet.

Keywords: Application of Artificial Intelligence, Neural Networks and their Applications, Machine Learning
Abstract: Accurately forecasting significant wave height (SWH) is meaningful since SWH is an essential parameter in coastal and ocean engineering. In order to accurately predict SWH, we propose the OSP-FEAN method, which optimizes significant wave height prediction by feature engineering and attention network. Specifically, we conduct feature engineering by adding the first-order to twelfth-order lag variables of SWH to the input set for feature enhancement and using the random forest algorithm for feature selection. Moreover, we construct a sequence to sequence neural network. In order to improve the forecast accuracy, we add an attention mechanism based on the memory layer to this neural network. Finally, extensive experiments with observed data at different stations are conducted to verify the effectiveness of our method on 6-h, 12-h and 24-h predictions, especially the superiority in outlier prediction.

Keywords: System Modeling and Control, Control of Uncertain Systems, Fault Monitoring and Diagnosis
Abstract: This paper presents a new approach to co-design of state feedback and dynamic proportional observer for a class of uncertain linear systems with disturbances and faults. Our objective is to simultaneously estimate the state, the fault and use them to control the system. An H_infty index is used in order to minimize the effect of the disturbance input on the state and fault estimation errors. The state feedback and observer gains are obtained from an LMI problem. The procedure is applied to the dynamics of a 2DOF helicopter. It shows its ability to well handle the system.

Keywords: System Modeling and Control, Service Systems and Organizations, Infrastructure Systems and Services
Abstract: Most business continuity planning methods help provide adequate precautions in case of a crisis situation to keep the organization’s main activities running uninterruptedly in the moment of the crisis. One way to accurately improve the rules that arise from business continuity planning (BCP) would be to (i) proceed with the data multidimensions and multi-sources (physical, societal, financial, etc.) through a purpose-driven Information System (IS), (ii) build customized integral robustness indicators and (iii) use them to qualify the business continuity strategy. On this basis, this work’s findings make firms, particularly healthcare organizations, less sensitive to catastrophes by designing an interactive dashboard that pilots the business continuity indicators and highlights decisive components and resources.

Keywords: System Modeling and Control, Modeling of Autonomous Systems
Abstract: Capacity limitation and communication delay in satellite communication are issues in the development of remote ship control system. This study proposes a system that receives ship’s motion data instead of images from a ship and reproduces the situation at sea in virtual reality. The results of experiments with an actual ship shows that it is possible to reproduce seamless images under small capacity and delay of communication and maneuver a ship in virtual reality.

Keywords: System Modeling and Control, Discrete Event Systems
Abstract: We recently presented the so-called allagmatic method, which includes a system metamodel providing a framework for describing, modelling, simulating, and interpreting complex systems. Its development and programming was guided by philosophy, especially by Gilbert Simondon's philosophy of individuation, Alfred North Whitehead’s philosophy of organism, and concepts from cybernetics. Here, a mathematical formalism is presented to better describe and define the system metamodel of the allagmatic method, thereby further generalising it and extending its reach to a more formal treatment and allowing more theoretical studies. By using the formalism, an example for such a further study is provided with mathematical definitions and proofs for model creation and equivalence of cellular automata and artificial neural networks.

Keywords: System Modeling and Control, Consumer and Industrial Applications, Decision Support Systems
Abstract: This paper proposes transferred initialization with modified fully connected layers for COVID-19 diagnosis. Convolutional neural networks (CNN) achieved a remarkable result in image classification. However, training a high-performing model is a very complicated and time-consuming process because of the complexity of image recognition applications. On the other hand, transfer learning is a relatively new learning method that has been employed in many sectors to achieve good performance with fewer computations. In this research, the PyTorch pre-trained models (VGG19_bn and WideResNet -101) are applied in the MNIST dataset for the first time as initialization and with modified fully connected layers. The employed PyTorch pre-trained models were previously trained in ImageNet. The proposed model is developed and verified in the Kaggle notebook, and it reached the outstanding accuracy of 99.77% without taking a huge computational time during the training process of the network. We also applied the same methodology to the SIIM-FISABIO-RSNA COVID-19 Detection dataset and achieved 80.01% accuracy. In contrast, the previous methods need a huge compactional time during the training process to reach a high-performing model. Codes are available at the following link: github.com/dipuk0506/SpinalNet

Keywords: Supervisory Control, Assistive Technology, Systems Safety and Security
Abstract: Anomaly detection technology relying on memory reconstruction leverages the difference in reconstruction errors between the normal and abnormal frames to achieve superior detection performance. However, there are still some challenges with this technology. First, the memory has insufficient representation capacity for features. Second, there is a contradiction between feature fusion and reconstruction. As feature fusion copies the abnormal patterns into the reconstructed frames, the abnormal frames are effectively reconstructed, reducing the detection performance. In response to these challenges, we use a memory update threshold to improve the representational power of memory. We also propose a dual-encoder anomaly detection model to restrict anomaly feature propagation. Experiment results demonstrate the effectiveness and robustness of our approach.

Keywords: Assistive Technology, Human-Computer Interaction, Human Factors
Abstract: Human Activity Recognition (HAR) is one of the fundamental building blocks of human assistive devices like orthoses and exoskeletons. There are different approaches to HAR depending on the application. Numerous studies have been focused on improving them by optimising input data or classification algorithms. However, most of these studies have been focused on applications like security and monitoring, smart devices, the internet of things, etc. On the other hand, HAR can help adjust and control wearable assistive devices, yet there has not been enough research facilitating its implementation. In this study, we propose several models to predict four activities from inertial sensors located in the ankle area of a lower-leg assistive device user. This choice is because they do not need to be attached to the user’s skin and can be directly implemented inside the control unit of the device. The proposed models are based on Artificial Neural Networks and could achieve up to 92.5% average classification accuracy.

Keywords: Multimedia Systems
Abstract: Natural image matting, since its born, has been regarded as a regression problem. Some previous works show the classification attributes in image matting are of great potential. Here we further explore the attributes to solve the matting problem in a pure classification way, which outperforms its regressive counterpart. We delved into the characteristics of α on alpha matte. Then we chop the continuous alpha value from 0 to 1 into 21 segments, each of which accounts for one class. The 21 classes are encoded by a simple number from 0 to 20 and each class represents a corresponding alpha value. Thus each pixel in image belongs to one class. Our model can be trained with both classification loss and regression loss. The latter loss is implemented with soft-argmax, which makes the selection of maximum index differentiable. Through softargmax, our model can be aware of the distance relationship among all classes. Moreover, we find this strategy enables the two kinds of losses to promote each other during training, which allows our model converge further.

Keywords: Assistive Technology, Human-Machine Interface, Brain-based Information Communications
Abstract: The paper presents a shared control architecture for non-holonomic mobile robots commanded through a motor imagery based Brain-Computer Interface (BCI). The overall system is aimed at assisting people to teleoperate a mobile robot in a simulated house-like scenario by resorting to two motor imagery commands. The developed architecture is structured in such a way that the user can drive the mobile robot while safety tasks, e.g. obstacle avoidance, are autonomously achieved, leaving complete autonomy to the mobile robot to let the latter adjust its configuration, e.g. aligning itself with a narrow passage. The overall architecture has been realized by developing control modules with the ROS environment, while the OpenVibe framework has been adopted to process the EEG signals. The effectiveness of the proposed architecture has been validated through experiments where a healthy user, wearing a Unicorn g.tec BCI, performs an assisted teleoperation task through motor imagery sessions with a Turtlebot robot.

Keywords: Human Factors, Assistive Technology, Design Methods
Abstract: Human-robot collaboration offers the advantage of combining human characteristics and robotic capabilities, balancing individual weaknesses. In the inclusive project Next Generation, we are exploring the possibility of using collaborative robots as assistive devices for people with severe and multiple disabilities. An important step in implementing an inclusive workstation with a collaborative robot is determining the level of assistance required. Therefore, we developed a novel methodology and a capability-based software tool to determine the individual level of challenge. In this paper, we present the developed tool and its validation. We validate the methodology and tool using an industrial sample application from first labor market incorporating participants with varying mental and physical disabilities.

Keywords: Passive BMIs, BMI Emerging Applications
Abstract: Many people experience motion sickness. In order to analyze a driver's motion sickness state and prevent accidents, a method of estimating the degree of motion sickness based on bio-signals is emerging. The brain-computer interface (BCI) systems using electroencephalogram (EEG) are used as the most direct method of estimating motion sickness conditions. However, EEG-based systems suffer from variability between subjects and over time, so a calibration process is required for every use. To address this problem, we mitigate the need for calibration through cross-subject transfer learning between the target data and the multi-subjects source data. All experiments were conducted in a domain adaptation setting. Meanwhile, we assume that there is an ordinal relationship between motion sickness scores. Thus, we performed an ordinal classification task so that the feature vectors were mapped by reflecting the ordinal characteristics according to the motion sickness state. In this paper, we propose a motion sickness classification BCI framework in combination with ordinal classification, resting-state prototype-based ordinal distance learning, and a subject-specific embedding module. Taking into account constraints of ordinal rank, the feature extractor is trained with prototype-based ordinal distance learning to measure the relative distance between the resting-state and motion sickness state. We further utilize an embedding module that encodes subject-specific information combined with task discriminative features to be effective for domain adaptation tasks. The proposed framework achieved the highest performance (accuracy 60.21 %) through comparative experiments with other models.

Keywords: Active BMIs, BMI Emerging Applications, Other Neurotechnology and Brain-Related Topics
Abstract: Brain-computer interface (BCI) is a technology that controls computers by reflecting users' intentions. Especially the electroencephalogram (EEG)-based BCI systems have been developed because of their potential utility. In BCI studies, controlling the drone swarm is one of the important issues since it improves work efficiency and safety. Also, current research has investigated how the drone swarms are controlled by imagining their formations using visual imagery (VI)-based EEG signals. The raw EEG signals and the spectrogram are widely used as input representations for decoding EEG signals. However, the decoding performance of the VI-based EEG signals is low to control the drone swarm due to noise in the raw EEG signals and information loss problems that may arise in the spectrogram. In this paper, we develop the EEGgram generator that extracts spectrogram-like features from the raw EEG signals minimizing information loss problems. Also, we propose the EEGgramNet, which could extract significant information from VI-based EEG signals using both the spectrogram and the EEGgram as inputs. The proposed method outperforms an accuracy of 0.643, which is 8.4 % higher than that of the best conventional method. Hence, we demonstrate the possibility of constructing a VI-based BCI system to control the drone swarm by imagining its formations.

Keywords: Human-Computer Interaction, Augmented Cognition, Human-Machine Cooperation and Systems
Abstract: Mental stress is a significant factor in the development of a wide variety of psychological, emotional, behavioral, and physical illnesses. It is critical to accurately quantify mental stress, which needs reliable neuroimaging to monitor stress levels. In this work, we used a modified Stroop Color Word Task (SCWT) with time constraints and negative feedback to elicit two distinct degrees of stress in the workplace. We then used salivary alpha amylase concurrently with functional near-infrared spectroscopy (fNIRS) to quantify the level of stress. We propose Long Short-Term Memory (LSTM) to decode the two classes of mental stress based on the fNIRS time series. We found that LSTM classified mental stress levels with an average accuracy of 72.5%. The findings indicated that the developed LSTM could be used to effectively classify mental stress using fNIRS time series.

Keywords: Passive BMIs, BMI Emerging Applications
Abstract: Drowsy driving causes severe road traffic accidents and significantly threatens road driving. Recently, electroencephalogram (EEG)-based drowsiness state classification has gained attention in the field of brain-computer interface (BCI). Because of the inter-and intra-subject variability of EEG signals, EEG-based drowsiness state classification is still challenging in developing an estimator applicable to unseen subjects. Generally, calibration sessions are required to tune the model with subject-specific data. In this paper, we propose an EEG-based driver drowsiness state (i.e., alert and drowsy) classification framework that improves the generalization performance to unseen subjects. Style features of multi-domain instances are mixed to generate unseen domains, and the distance of labels within classes is minimized to learn robust representations. Experiments were conducted on EEG data acquired from a drowsy driving experiment in a simulated-driving environment. Our proposed framework achieved an accuracy of 77.26%, an F1-score of 0.6266, and a recall of 0.6813 across eleven subjects in leave-one-subject-out cross-validation. The experimental results showed an improvement in the generalization performance for novel target subjects in driver drowsiness state classification and demonstrated the potential for calibration-free BCI.

Keywords: Active BMIs, Passive BMIs, BMI Emerging Applications
Abstract: Brain-Computer Interface (BCI) systems have a great impact on improving people's lives. One of the popular BCI implementations is the Event-Related Potential (ERP)-based spelling system which decodes electroencephalogram (EEG) signals to identify a target character. The effectiveness of BCI systems highly depends on the single trial decoding accuracy; however, the EEG signals are contaminated with diverse artifacts which leads to a poor signal-to-noise ratio. Therefore, various filtering algorithms (e.g., FFT, CSP, Laplacian, PCA) have been applied to find the optimal subset of feature spaces in the temporal and spatial domains. These preprocessing steps could efficiently discard the artifacts and have shown superior performance with typical linear classifiers. However, there is a risk that the informative subspace can be also eliminated by the unsupervised learning process, and this algorithm is not proper to be employed in the end-to-end deep-learning architectures where all modules are differentiable. This study aims to propose a generalized deep neural network model by denoising the ERP signals and initializing the Convolutional Neural Network (CNN) model parameters based on the autoencoder. Proposed CNN models indicate - 98.2% spelling performance and - 91.5% single trial accuracy which outperformed the state-of-the-art CNN models.

Keywords: System Architecture
Abstract: In recent years, with the increasing demand for edge AI, the need for faster AI inference on embedded CPUs is growing. As a result, researchers have focused on various redundancies in neural networks and proposed network acceleration methods. The basic element of most neural networks is a combination of linear transformations and nonlinear activations. We believe this linearity itself is redundant and propose a new method for designing a fast neural network by using nonlinear b-pline functions instead of linear multiplications. We propose an end-to-end design flow for high-speed b-spline neural networks, which shows an actual speedup of 2.4 to 4.1 times compared to conventional neural networks when the network is implemented on Raspberry Pi Zero and Raspberry Pi Pico.

Keywords: System Architecture, Discrete Event Systems, System Modeling and Control
Abstract: A balanced method for the merging and splitting transformation procedures for incompletely specified finite state machines implemented on field programmable logic arrays is proposed. In this method, such optimization criteria as the power consumption, speed of operation and cost of implementation are considered already in the early phase of FSM synthesis. The method also takes into account the technological features of programmable logic and the state assignment method. The transformation quality ratio is calculated on the base of estimations of power, speed and area parameters and the user can select the direction of the optimization by setting weights for each criterion. The approach to the estimation of optimization criteria values is presented and experimental results are also discussed.

Keywords: System Architecture
Abstract: Dynamic complex environment requires resilient system-of-systems (SoS) architecture that can effectively deal with uncertainty. However, increasing resilience could affect other evaluation metrics, such as cost and effectiveness. To simplify the trade-off analysis process in the early design phase, this paper explores the use of a representative index that directly illustrates a balanced range of effectiveness, cost, and resilience. Specifically, this paper compares the combat SoS to an ecological network and develops a functional dependency network based effectiveness, resilience, and order-degree analysis method. Compared with the previous order-degree research, the proposed combat SoS architecture modeling and calcuation based on fucntional dependency network can reflect the rules of data exchanging more realistically. Based on the application to a notional anti-ship combat SoS, the order-degree is able to reflect the relationship between SoS resilience index and cost-effectiveness index (CEI) in the given scenarios. This implies that the order-degree has a good opportunity to serve as an indicator for supporting balanced resilient SoS design.

Keywords: System Architecture, Infrastructure Systems and Services, Large-Scale System of Systems
Abstract: Profile-guided optimization (PGO) remains one of the most popular optimization strategies in code generation optimization. Function reordering is an essential step for profile-guided optimization. The state-of-art function reordering method performs on a unidirectional function call graph where nodes and edges define functions and caller-callee pairs. Each edge is labeled by its call frequency. However, we demonstrate that a bidirectional function call graph can represent the memory call better. We use a reinforcement learning algorithm SARSA to choose the appropriate order of functions by maximizing the total numbers of the function call through a bidirectional function call graph. In this paper, we use a self-developed tool to reordering functions. We first illustrate how our RL-based algorithm generates a new function order. Then we evaluate three algorithms on various applications, including Redis, Protobuf, and SPEC CPU benchmark. Our experiment results indicate that the new algorithm outperforms the other two algorithms in various applications, improving the resulting performance of practical applications. Especially on Redis, the performance is improved by 4.2% on SARSA, which is better than C3 (3.4%) and ph (2.8%).

Keywords: Communications, Smart Sensor Networks, Discrete Event Systems
Abstract: Pocket Switched Networks (PSNs) are an evolution of mobile ad-hoc and Delay Tolerant Networks in which there is no assumption made about the existence of a complete path between two nodes wishing to communicate, thus making routing even more challenging. Various routing algorithms have been proposed over the years like Epidemic, Spray and Wait, ProPHET, etc. There is a separate class of routing algorithms that exploit social structuralism to selectively forward messages to the best candidates. One such algorithm is BUBBLE Rap, which takes into account the notion of popularity and communities to make forwarding decisions. However, popularity as a sole deal breaker is a rigid policy. We propose a decentralized community convergence-based message forwarding approach viz., De-COP that makes use of the familiarity metric which is dictated by the characteristic of a community to make message forwarding decisions. When familiarity is used for forwarding messages in converged communities, the messages are delivered with low latency and high probability. A community is defined as converged when the change in its membership is gradual. The forwarding node also takes into account the buffer availability at the target node to reduce delay and message loss probability. The results were obtained using ONE simulator on two popular datasets, i.e., Infocomm06 and Cambridge to demonstrate the efficacy of the proposed De-COP approach in terms of improved delivery probability and message overhead ratio compared to ProPHET and BUBBLE Rap.

Keywords: Deep Learning, Application of Artificial Intelligence, Computational Intelligence
Abstract: Deep learning has witnessed remarkable success in many application domains and is now shifting towards training super deep models with extremely large scale labeled or unlabeled data on expensive computational resources. In this talk, I will present some of the recent progress. Specifically, I will first show the PAC-Bayes generalization bounds and present some practical implications for new algorithm designs. Then, I will propose an efficient architecture design for visual transformers, named ViTAE, by exploring the intrinsic inductive biases. Next, he will introduce a novel self-supervised training method called RegionCL, which uses a simple region swapping strategy to build effective supervisory signals from rich positive/negative pairs at both the instance level and the region level. It greatly advances the ability of representative self-supervised leaning frameworks including MoCo, SimCLR, and SimSam. Finally, some promising applications of visual transformers and self-supervised leaning will be presented, including image classification, object detection, semantic segmentation, and pose estimation.

Keywords: Intelligent Power Grid, Distributed Intelligent Systems, Communications
Abstract: The digital transformation of our power system leads to lots of benefits: more insight, more flexibility, more preparedness for an uncertain and dynamic future. Equipping all power system assets with digital functions, however, also imports phenomena from the ICT world: complexity, software bugs, race- conditions, interoperability questions, and - last but not least - cyber-security problems. Activist hackers, terrorists, digital vandals, state player attacks: they all can aim at the digital assets of modern power systems in order to impact the physical half of it. Industrial control systems for power systems such as IEC 61850 and the smart IoT- enabled grid-edge are elements of the so-called attack surface. This talk will introduce you to cyber-physical power systems, show some modeling options, and explain which threats we have to deal with now and in future. It will also explain some known attacks of recent years and demonstrate you how to hack digital substation protection leading to cascading outages - more prominently known as blackouts.

Keywords: Technology Assessment
Abstract: Content

PhD: Why and How?

Embarking on doing a Ph. D. is an important decision. It should not be taken lightly. In this presentation, the difficulties one is likely to meet on the way to a doctoral degree are discussed and if you cannot come up with strong arguments in favor of it, it is suggested that you should perhaps go for a master’s degree. The second part of the presentation focuses on the journey itself. The skills that you should have or need to acquire for a successful completion of your journey as sanely as possible are discussed. The final remark is that dissertations are never completed but abandoned.

The Art of Technical Paper Writing

Now that you have embarked on your Ph. D. journey, you will be expected to publish. It may even be a prerequisite to your degree. In this presentation, the art of technical paper writing is elaborated upon. Firstly a three dimensional view of dissemination of research results is introduced and where a good scientific paper should position itself is discussed. Secondly, the basic rules that a good scientific paper needs to follow are presented in relation with the typical arguments behind a quick rejection. Thirdly, how to organize a technical paper is discussed with specific references to its main parts, i.e. the title, the abstract, the body, the mathematics used, the conclusions and the references.

Keywords: Human-Machine Cooperation and Systems, Human-Computer Interaction, Human-Machine Interface
Abstract: Content

- IEEE Press/Wiley Book Series on Human-Machine Systems, Giancarlo Fortino
- IEEE Press/Wiley Book Series on Systems Science and Engineering, MengChu Zhou

Keywords: Technology Assessment
Abstract: Meeting of the IEEE SMC Technical Committee Chairs with the IEEE SMC VPs for Cybernetics, Human-Machine Systems, and Systems Science and Engineering

Keywords: Technology Assessment
Abstract: Annual meeting of the IEEE SMC Standards Committee (IEEE SA SMC/SC)

Keywords: Passive BMIs, Active BMIs, BMI Emerging Applications
Abstract: Further details are provided at https://musaelab.ca/bmi22/smc2022_bmi.html and https://www.br41n.io/IEEE-SMC-2022