We have used our IoT platform selection framework to select a suitable IoT platform for Industrial Energy Management Systems

Abstract— One of the most prevalent diseases, skin cancer, has been proven to be treatable at an early stage. Thus, techniques that allow individuals to identify skin cancer symptoms early are in great demand. This paper proposed an interactive skin lesion diagnosis system based on the ensemble of multiple sophisticated CNN models for image classification. The performance of ResNet50, ResNeXt50, ResNeXt101, EfficientNetB4, Mobile-NetV2, MobileNetV3, and MnasNet are investigated separately as ensemble components. Then, using various criteria, we constructed ensembles and compared the accuracy they achieved. Moreover, we designed a method to update the ensemble for new data and examined its performance. In addition, a few natural language processing (NLP) techniques were used to make our system more user-friendly. To integrate all the functionalities, we built a user interface with PyQt5. As a result, MobileNetV3 achieved 91.02% as the best accuracy among all single models; ensemble weighted by cubic precision achieved 92.84% accuracy as the highest one in this study; a notable improvement in accuracy demonstrated the effectiveness of the model updating approach, and a system with all of the desired features was successfully developed. These findings benefit in two aspects. For model performance, applying cubic precisions can increase ensemble learning classification accuracy. For the developed diagnosis system, it can aid in the

based on the new discovery of moving photon do create force and a formula of describe this new discovery: Applying this discovery from the micro world to the macro world, show that from atomic world to the galaxy world, the nature has been working obeys this law, and their actions can all be described by this formula. The Coulomb law and Newton law of university gravitation are only calculation formula in the specific condition status.

In this paper, we present the design and development of MIDAS. MIDAS is intended to be a portable immersive system inclusive of an exoskeleton, a smell releasing device, and a VR game that will help stroke patients feel more motivated during a hand rehabilitation session. MIDAS engages four out of five senses during rehabilitation. MIDAS uses EMG signals taken from the subject’s forearm to predict their intention before activating physical assistance (tactile) in the opening or closing of the fingers of their hand. To evaluate the performance of MIDAS, a pilot study is done using established tools to quantify the user experience, comfort, pleasure, and motivation of the subjects in using MIDAS for rehabilitation exercises.

in this study we have focus the the importance of using IoT , Big Data and Edge computing in the information processing of smart grid. how these technologies are used to increased the efficiency of information processing in smart grid

In the last decade, the increase in the number of active and passive earth observation satellites has provided us with more remote sensing data. This fact has led to increased interests in the field of fusion of the different satellite data since some of the satellites have properties complementary to one another. Fusion techniques can improve the estimation in areas of interest (AOIs) by using complementary information and inferring unknown parameters. However, when the observation area is large, extensive human labor and domain expertise are required for processing and analysis. Thus, we propose a neural network which combines and analyzes the data obtained from synthetic aperture radars (SAR) and optical sensors. The neural network employs a modified logarithmic activation function, unlike conventional networks, to realize inverse mapping for significant feature analysis based on dynamics consistent with its forward processing. In this paper, we focus on earthquake damage detection by dealing with the data of the 2018 Sulawesi earthquake in Indonesia. The fusion-based results show increased classification accuracy compared to the results of independent sensors. We further attempt to understand which input features are the significant contributors for which classification outputs by inverse-mapping in the data fusion neural network. We observe that inverse mapping shows reasonable explanations in a consistent manner. It also indicates contributions of features different from straightforward counterparts, namely pre- and post-seismic features, in the detection of particular classes.

With the acquisition of massive condition monitoring data, how to realize real-time and efficient intelligent fault diagnosis is the focus of current research. Inspired by the ideas of compressed sensing (CS) and deep extreme learning machines (DELM), a data-driven lightweight framework is proposed to accelerate intelligent fault diagnosis. The integrated framework contains two modules: data sampling and fault diagnosis. Data sampling module projects the intensive original monitoring data into lightweight compressed sampling data non-linearly, which can effectively reduce the pressure of transmission, storage and calculation. Fault diagnosis module digs deeply into the inner connection between the compressed sampled signal and the fault types to realize accurate fault diagnosis. This work has three meaningful points. First, we believe that the bearing vibration signal is not strictly sparse in the transform domain. Second, we verified that the sparse signal after compressed sampling can be directly used for fault diagnosis without being reconstructed. Third, adding a kernel function to the DELM can perfectly map the low-dimensional inseparable features after compressed sampling to the high-dimensional space non-linearly to make it linearly separable and thus improve the classification accuracy

Shaped and multiple beams using planar phased arrays. This paper presents a new synthesis and computation method to generate user-specified multiple beams and shaped beams in any arbitrary 3D space. The computation method can generate independently controllable simultaneous multiple beams with arbitrary peak powers. This method is extended to generate arbitrarily shaped beams, using a combination of optimally placed multiple beams at appropriate locations with specific power ratios.

It is widely believed that quantum computing is on the threshold of practicality, with performance that will soon greatly surpass that of classical computing. On the contrary, I argue that quantum computing does not currently exist, and probably never will. First, although quantum annealing systems have been demonstrated to solve practical optimization problems, they are actually performing classical analog annealing, with no quantum enhancement. In contrast, while systems of quantum gate arrays, which are expected to perform digital quantum computing, have been fabricated with up to ~ 100 qubits in several technologies, they have not performed any practical computations. This is not merely a question of excess noise; the theory of massive quantum entanglement, necessary for the desired performance, has never been actually been verified. The well-established quantum results such as electronic energy bands do not incorporate quantum entanglement. I suggest that the experimental observations in multi-qubit systems may be explained as the result of delocalized coupled oscillator modes, similar to that in electronic energy bands. Such coupled modes would not yield the exponential increase in degrees of freedom needed for quantum speedup, and hence would not be useful for computing. Tests on these multi-qubit systems should be able to distinguish these two models. The quantum computing research community really needs to address this issue.

Underwater sensor networks have become increasingly interesting in the past four decades. They can be used in a multitude of scenarios, commercial and military alike. Underwater networks can communicate in several ways, but when nodes are far apart, underwater acoustic communication is the only feasible way. The complex underwater acoustic channel puts high demands on the network protocols. The physical layer needs to contend with short coherence times, high intersymbol interference and significant Doppler spread. The routing protocol needs to handle intermittent connectivity and mobile network topologies, such as autonomous underwater vehicle networks. The medium access control protocol needs to manage medium access with high latency and potentially high packet loss ratios without congesting the network. The available acoustic modems are still rather expensive, which limits the size of a sensor network. Voices have also been raised from the academia for a paradigm shift, from hardware-defined, proprietary modems to software-defined, open-architecture modems, in order to accelerate research in the field and enable interoperability. This paper reviews the recent advancements in designing and implementing underwater networks on several levels and discusses some interesting approaches to underwater ad-hoc networking. The focus lies on acoustic communication.

This paper reports a new device for electrocardiogram (ECG) signal monitoring and software for signal analysis and artificial intelligence (AI) assisted diagnosis. The hardware mitigates the signal loss common in previous products by enhancing the ergonomy, flexibility, and battery life. The power efficiency is optimized by design using switching converters, ultra-low-power components, and efficient signal processing. It enables 14-day of uninterrupted ECG monitoring and connectivity with a smartphone and microSD card storage. The software is implemented in Android app and web-based platforms via Internet of Things (IoT). This component provides cloud-based and local storage and uses AI for arrhythmia detection. The arrhythmia detection algorithm shows 98.7% accuracy using Artificial Neural Network and K-Nearest Neighbors methods, and 98.1% using Decision Tree method on test data set.

This work presents DeepLSF, a framework for time series forecasting that fuses knowledge driven techniques with data driven neural networks. The proposed framework achieves State-Of-The-Art results on three different real world time series forecasting datasets.

Face detection and subsequent localization of facial landmarks are the primary steps in many face applications. Numerous algorithms and benchmark datasets have been introduced to develop robust models for the visible domain. However, varying conditions of illumination still pose challenging problems. In this regard, thermal cameras are employed to address this problem, because they operate on longer wavelengths. However, thermal face and facial landmark detection in the wild is an open research problem because most of the existing thermal datasets were collected in controlled environments. In addition, many of them were not annotated with face bounding boxes and facial landmarks. In this work, we present a thermal face dataset with manually labeled bounding boxes and facial landmarks to address these problems. The dataset contains 9,982 images of 147 subjects collected under controlled and uncontrolled conditions. As a baseline, we trained the YOLOv5 object detection model and its adaptation for face detection, YOLO5Face, on our dataset. In addition to our test set, we evaluated the models on the external RWTH-Aachen thermal face dataset to show the efficacy of our dataset. We have made the dataset, source code, and pre-trained models publicly available at https://github.com/IS2AI/TFW to bolster research in thermal face analysis.

Power systems with high penetration of renewable energy contain various uncertainties. Scenario-based optimization problems need a large number of discrete scenarios to obtain a reliable approximation for the probabilistic model. It is important to choose typical scenarios and ease the computational burden. This paper presents a scenario reduction network model based on Wasserstein distance. Entropy regularization is used to transform the scenario reduction problem into an unconstrained problem. Through an explicit neural network structure design, the output of the scenario reduction network corresponds to Sinkhorn distance function. The scenario reduction network can generate the typical scenario set through unsupervised learning training. An efficient algorithm is proposed for continuous/discrete scenario reduction. The superiority of the scenario reduction network model is verified through case studies. The numerical results highlight high accuracy and computational efficiency of the proposed model over state-of-the-art model making it an ideal candidate for large-scale scenario reduction problems

We describe a simple and practical algorithm for compact routing on graphs which admit compact and balanced vertex separators. Using a recursive nested dissection of then-vertex graph based on these separators, we construct routing tables with as few as O(log n) entries per vertex in a preprocessing step. They support handshaking-based routing on the graph with moderate stretch, where the handshaking can be implemented similarly to a DNS lookup. We describe a basic version of the algorithm that requires modifiable headers and a more advanced version which eliminates this need and provides better stretch. A number of algorithmic parameters control a graceful tradeoff between the size of the routing tables and the stretch. Our routing algorithm is most effective on planar graphs and unit disk graphs of moderate edge/vertex density.

In this paper, a varying-gain zeroing (or Zhang) neural network (VG-ZNN) is proposed to obtain the online solution of the time-varying linear equation and inequality system. Distinguished from the fixed-value design parameter in the original zeroing (or Zhang) neural network (ZNN) models, the design parameter of the VG-ZNN model is a nonlinear function that changes with time. The VG-ZNN model composed of the new time-varying design parameter we proposed can achieve fixed-time convergence and tolerate time-varying bounded noise and time-varying derivable noise. The theoretical detailed analysis of the convergence and robustness of the VG-ZNN model are given.

To obtain an accurate digital surface model of the small watershed topography of a forested area while reducing time and labor costs, we used a consumer-grade unmanned aerial vehicle (UAV) with a build-in real-time kinematic global navigation satellite system. The applicability of structure-from-motion (SfM) multi-view stereo processing with post-processing kinematic (PPK) correction of the positional coordinate data (the UAV-PPK-SfM method) was tested. Nine verification points were set up in a small (0.5 km2) watershed, based on a check dam in the headwaters of a forest area. The location information of the verification points extracted from the digital surface model acquired by UAV-PPK-SfM and the overall working time were compared with the corresponding location information and working time of a traditional field survey using a total station. The results showed that the vertical error between the total station and each verification point at an altitude of 150 m ranged from 0.006 to 0.181 m. The working time of the UAV-PK-SfM survey was 10 % of that of the total station survey (30 min). The UAV-PPK-SfM workflow proposed in this study shows that wide-area, non-destructive topographic surveying, including fluvial geomorphological mapping, is possible with a vertical error of ±0.2 m in small watersheds (<0.5 km2). This method will be useful for rapid topographic surveying in inaccessible areas during disasters, such as monitoring debris flow at check dam sites, and for efficient topographic mapping of steep valleys in forested areas where the positioning of ground control points is a laborious task.

Trust and security of microelectronic systems are a major driver for game-changing trends like autonomous driving or the internet of things. These trends are endangered by threats like soft- and hardware attacks or IP tampering – wherein often hardware reverse engineering (RE) is involved for efficient attack planning. The constant publication of new RE-related scenarios and countermeasures renders a profound rating of these extremely difficult. Researchers and practitioners have no tools or framework which aid a common, consistent classification of these scenarios. In this work, this rating framework is introduced: the common reverse engineering scoring system (CRESS). The framework allows a general classification of published settings and renders them comparable. We introduce three metrics: exploitability, impact, and a timestamp. For these metrics, attributes are defined which allow a granular assessment of RE on the one hand, and attack requirements, consequences, and potential remediation strategies on the other. The system is demonstrated in detail via five case studies and common implications are discussed. We anticipate CRESS to evaluate possible vulnerabilities and to safeguard targets more proactively.