Search

An electrocardiogram (ECG) is an essential piece of medical equipment that helps diagnose various heart-related conditions in patients. An automated diagnostic tool is required to detect significant episodes in long-term ECG records. It is a very challenging task for cardiologists to analyze long-term ECG records in a short time. Therefore, a computer-based diagnosis tool is required to identify crucial episodes. Myocardial infarction (MI) and conduction disorders (CDs), sometimes known as heart blocks, are medical diseases that occur when a coronary artery becomes fully or suddenly stopped or when blood flow in these arteries slows dramatically. As a result, several researchers have utilized deep learning methods for MI and CD detection. However, there are one or more of the following challenges when using deep learning algorithms: (i) struggles with real-life data, (ii) the time after the training phase also requires high processing power, (iii) they are very computationally expensive, requiring large amounts of memory and computational resources, and it is not easy to transfer them to other problems, (iv) they are hard to describe and are not completely understood (black box), and (v) most of the literature is based on the MIT-BIH or PTB databases, which do not cover most of the crucial arrhythmias. This paper proposes a new deep learning approach based on machine learning for detecting MI and CDs using large PTB-XL ECG data. First, all challenging issues of these heart signals have been considered, as the signal data are from different datasets and the data are filtered. After that, the MI and CD signals are fed to the deep learning model to extract the deep features. In addition, a new custom activation function is proposed, which has fast convergence to the regular activation functions. Later, these features are fed to an external classifier, such as a support vector machine (SVM), for detection. The efficiency of the proposed method is demonstrated by the experimental findings, which show that it improves satisfactorily with an overall accuracy of 99.20% when using a CNN for extracting the features with an SVM classifier. [ABSTRACT FROM AUTHOR]

Mechanical equipment failure may cause massive economic and even life loss. Therefore, the diagnosis of the failures of machine parts in time is crucial. The rolling bearings are one of the most valuable parts, which have attracted the focus of fault diagnosis. Many successful rolling bearing fault diagnoses have been made based on machine learning and deep learning. However, most diagnosis methods still rely on complex signal processing and artificial features, bringing many costs to the deployment and migration of diagnostic models. This paper proposes an end-to-end continuous/discontinuous feature fusion method for rolling bearing fault diagnosis (C/D-FUSA). This method comprises long short-term memory (LSTM), convolutional neural networks (CNN) and attention mechanism, which automatically extracts the continuous and discontinuous features from vibration signals for fault diagnosis. We also propose a contextual-dependent attention module for the LSTM layers. We compare the method with the other simpler deep learning methods and state-of-the-art methods in rolling bearing fault data sets with different sample rates. The results show that our method is more accurate than the other methods with real-time inference. It is also easy to be deployed and trained in a new environment. [ABSTRACT FROM AUTHOR]

Machine learning (ML) is a key technology in smart manufacturing as it provides insights into complex processes without requiring deep domain expertise. This work deals with deep learning algorithms to determine a 3D reconstruction from a single 2D grayscale image. The potential of 3D reconstruction can be used for quality control because the height values contain relevant information that is not visible in 2D data. Instead of 3D scans, estimated depth maps based on a 2D input image can be used with the advantage of a simple setup and a short recording time. Determining a 3D reconstruction from a single input image is a difficult task for which many algorithms and methods have been proposed in the past decades. In this work, three deep learning methods, namely stacked autoencoder (SAE), generative adversarial networks (GANs) and U-Nets are investigated, evaluated and compared for 3D reconstruction from a 2D grayscale image of laser-welded components. In this work, different variants of GANs are tested, with the conclusion that Wasserstein GANs (WGANs) are the most robust approach among them. To the best of our knowledge, the present paper considers for the first time the U-Net, which achieves outstanding results in semantic segmentation, in the context of 3D reconstruction tasks. Unlike the U-Net, which uses standard convolutions, the stacked dilated U-Net (SDU-Net) applies stacked dilated convolutions. Of all the 3D reconstruction approaches considered in this work, the SDU-Net shows the best performance, not only in terms of evaluation metrics but also in terms of computation time. Due to the comparably small number of trainable parameters and the suitability of the architecture for strong data augmentation, a robust model can be generated with only a few training data. [ABSTRACT FROM AUTHOR]

Drones have attracted extensive attention for their environmental, civil, and military applications. Because of their low cost and flexibility in deployment, drones with communication capabilities are expected to play key important roles in Fifth Generation (5G), Sixth Generation (6G) mobile networks, and beyond. 6G and 5G are intended to be a full-coverage network capable of providing ubiquitous connections for space, air, ground, and underwater applications. Drones can provide airborne communication in a variety of cases, including as Aerial Base Stations (ABSs) for ground users, relays to link isolated nodes, and mobile users in wireless networks. However, variables such as the drone's free-space propagation behavior at high altitudes and its exposure to antenna sidelobes can contribute to radio environment alterations. These differences may render existing mobility models and techniques as inefficient for connected drone applications. Therefore, drone connections may experience significant issues due to limited power, packet loss, high network congestion, and/or high movement speeds. More issues, such as frequent handovers, may emerge due to erroneous transmissions from limited coverage areas in drone networks. Therefore, the deployments of drones in future mobile networks, including 5G and 6G networks, will face a critical technical issue related to mobility and handover processes due to the main differences in drones' characterizations. Therefore, drone networks require more efficient mobility and handover techniques to continuously maintain stable and reliable connection. More advanced mobility techniques and system reconfiguration are essential, in addition to an alternative framework to handle data transmission. This paper reviews numerous studies on handover management for connected drones in mobile communication networks. The work contributes to providing a more focused review of drone networks, mobility management for drones, and related works in the literature. The main challenges facing the implementation of connected drones are highlighted, especially those related to mobility management, in more detail. The analysis and discussion of this study indicates that, by adopting intelligent handover schemes that utilizing machine learning, deep learning, and automatic robust processes, the handover problems and related issues can be reduced significantly as compared to traditional techniques. [ABSTRACT FROM AUTHOR]

It is important to reduce the danger of collecting flame image data sets by compositing flame images by computer. In this paper, a Global-Local mask Generative Adversarial Network (FGL-GAN) is proposed to address the current status of low quality composite flame images. First, FGL-GAN adopts a hierarchical Global-Local generator structure, to locally render high-quality flame halo and reflection, while also maintaining a consistent global style. Second, FGL-GAN incorporates the fire mask as part of the input of the generation module, which improves the rendering quality of flame halo and reflection. A new data augmentation technique for flame image compositing is used in the network training process to reconstruct the background and reduce the influence of distractors on the network. Finally, FGL-GAN introduces the idea of contrastive learning to speed up network fitting and reduce blurriness in composite images. Comparative experiments show that the images composited by FGL-GAN have achieved better performance in qualitative and quantitative evaluation than mainstream GAN. Ablation study shows the effectiveness of the hierarchical Global-Local generator structure, fire mask, data augmentation, and MONCE loss of FGL-GAN. Therefore, a large number of new flame images can be composited by FGL-GAN, which can provide extensive test data for fire detection equipment, based on deep learning algorithms. [ABSTRACT FROM AUTHOR]

Modern agriculture incorporated a portfolio of technologies to meet the current demand for agricultural food production, in terms of both quality and quantity. In this technology-driven farming era, this portfolio of technologies has aided farmers to overcome many of the challenges associated with their farming activities by enabling precise and timely decision making on the basis of data that are observed and subsequently converged. In this regard, Artificial Intelligence (AI) holds a key place, whereby it can assist key stakeholders in making precise decisions regarding the conditions on their farms. Machine Learning (ML), which is a branch of AI, enables systems to learn and improve from their experience without explicitly being programmed, by imitating intelligent behavior in solving tasks in a manner that requires low computational power. For the time being, ML is involved in a variety of aspects of farming, assisting ranchers in making smarter decisions on the basis of the observed data. In this study, we provide an overview of AI-driven precision farming/agriculture with related work and then propose a novel cloud-based ML-powered crop recommendation platform to assist farmers in deciding which crops need to be harvested based on a variety of known parameters. Moreover, in this paper, we compare five predictive ML algorithms—K-Nearest Neighbors (KNN), Decision Tree (DT), Random Forest (RF), Extreme Gradient Boosting (XGBoost) and Support Vector Machine (SVM)—to identify the best-performing ML algorithm on which to build our recommendation platform as a cloud-based service with the intention of offering precision farming solutions that are free and open source, as will lead to the growth and adoption of precision farming solutions in the long run. [ABSTRACT FROM AUTHOR]

The rapid growth of mobile data traffic will lead to the deployment of Ultra–Dense Networks (UDN) in the near future. Various networks must overlap to meet the massive demands of mobile data traffic, causing an increase in the number of handover scenarios. This will subsequently affect the connectivity, stability, and reliability of communication between mobile and serving networks. The inclusion of Unmanned Aerial Vehicles (UAVs)—based networks will create more complex challenges due to different mobility characterizations. For example, UAVs move in three–dimensions (3D), with dominant of line–of–sight communication links and faster mobility speed scenarios. Assuring steady, stable, and reliable communication during UAVs mobility will be a major problem in future mobile networks. Therefore, this study provides an overview on mobility (handover) management for connected UAVs in future mobile networks, including 5G, 6G, and satellite networks. It provides a brief overview on the most recent solutions that have focused on addressing mobility management problems for UAVs. At the same time, this paper extracts, highlights, and discusses the mobility management difficulties and future research directions for UAVs and UAV mobility. This study serves as a part of the foundation for upcoming research related to mobility management for UAVs since it reviews the relevant knowledge, defines existing problems, and presents the latest research outcomes. It further clarifies handover management of UAVs and highlights the concerns that must be solved in future networks. [ABSTRACT FROM AUTHOR]

Risky driving behavior seriously affects the driver's ability to react, execute and judge, which is one of the major causes of traffic accidents. The timely and accurate identification of the driving status of drivers is particularly important, since drivers can quickly adjust their driving status to avoid safety accidents. In order to further improve the identification accuracy, this paper proposes a risky-driving image-recognition system based on the visual attention mechanism and deep-learning technology to identify four types of driving status images including normal driving, driving while smoking, driving while drinking and driving while talking. With reference to ResNet, we build four deep-learning models with different depths and embed the proposed visual attention blocks into the image-classification model. The experimental results indicate that the classification accuracy of the ResNet models with lower depth can exceed the ResNet models with higher depth by embedding the visual attention modules, while there is no significant change in model complexity, which could improve the model recognition accuracy without reducing the recognition efficiency. [ABSTRACT FROM AUTHOR]

Fitness is important in people's lives. Good fitness habits can improve cardiopulmonary capacity, increase concentration, prevent obesity, and effectively reduce the risk of death. Home fitness does not require large equipment but uses dumbbells, yoga mats, and horizontal bars to complete fitness exercises and can effectively avoid contact with people, so it is deeply loved by people. People who work out at home use social media to obtain fitness knowledge, but learning ability is limited. Incomplete fitness is likely to lead to injury, and a cheap, timely, and accurate fitness detection system can reduce the risk of fitness injuries and can effectively improve people's fitness awareness. In the past, many studies have engaged in the detection of fitness movements, among which the detection of fitness movements based on wearable devices, body nodes, and image deep learning has achieved better performance. However, a wearable device cannot detect a variety of fitness movements, may hinder the exercise of the fitness user, and has a high cost. Both body-node-based and image-deep-learning-based methods have lower costs, but each has some drawbacks. Therefore, this paper used a method based on deep transfer learning to establish a fitness database. After that, a deep neural network was trained to detect the type and completeness of fitness movements. We used Yolov4 and Mediapipe to instantly detect fitness movements and stored the 1D fitness signal of movement to build a database. Finally, MLP was used to classify the 1D signal waveform of fitness. In the performance of the classification of fitness movement types, the mAP was 99.71%, accuracy was 98.56%, precision was 97.9%, recall was 98.56%, and the F1-score was 98.23%, which is quite a high performance. In the performance of fitness movement completeness classification, accuracy was 92.84%, precision was 92.85, recall was 92.84%, and the F1-score was 92.83%. The average FPS in detection was 17.5. Experimental results show that our method achieves higher accuracy compared to other methods. [ABSTRACT FROM AUTHOR]

Remaining useful life prediction is one of the essential processes for machine system prognostics and health management. Although there are many new approaches based on deep learning for remaining useful life prediction emerging in recent years, these methods still have the following weaknesses: (1) The correlation between the information collected by each sensor and the remaining useful life of the machinery is not sufficiently considered. (2) The accuracy of deep learning algorithms for remaining useful life prediction is low due to the high noise, over-dimensionality, and non-linear signals generated during the operation of complex systems. To overcome the above weaknesses, a general deep long short memory network-based approach for mechanical remaining useful life prediction is proposed in this paper. Firstly, a two-step maximum information coefficient method was built to calculate the correlation between the sensor data and the remaining useful life. Secondly, the kernel principal component analysis with a simple moving average method was designed to eliminate noise, reduce dimensionality, and extract nonlinear features. Finally, a deep long short memory network-based deep learning method is presented to predict remaining useful life. The efficiency of the proposed method for remaining useful life prediction of a nonlinear degradation process is demonstrated by a test case of NASA's commercial modular aero-propulsion system simulation data. The experimental results also show that the proposed method has better prediction accuracy than other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Computed tomography (CT) images play an important role due to effectiveness and accessibility, however, motion artifacts may obscure or simulate pathology and dramatically degrade the diagnosis accuracy. In recent years, convolutional neural networks (CNNs) have achieved state-of-the-art performance in medical imaging due to the powerful learning ability with the help of the advanced hardware technology. Unfortunately, CNNs have significant overhead on memory usage and computational resources and are labeled 'black-box' by scholars for their complex underlying structures. To this end, an interpretable graph-based method has been proposed for motion artifacts detection from head CT images in this paper. From a topological perspective, the artifacts detection problem has been reformulated as a complex network classification problem based on the network topological characteristics of the corresponding complex networks. A motion artifacts detection method based on complex networks (MADM-CN) has been proposed. Firstly, the graph of each CT image is constructed based on the theory of complex networks. Secondly, slice-to-slice relationship has been explored by multiple graph construction. In addition, network topological characteristics are investigated locally and globally, consistent topological characteristics including average degree, average clustering coefficient have been utilized for classification. The experimental results have demonstrated that the proposed MADM-CN has achieved better performance over conventional machine learning and deep learning methods on a real CT dataset, reaching up to 98% of the accuracy and 97% of the sensitivity. [ABSTRACT FROM AUTHOR]

Cyberattacks in the Internet of Things (IoT) are growing exponentially, especially zero-day attacks mostly driven by security weaknesses on IoT networks. Traditional intrusion detection systems (IDSs) adopted machine learning (ML), especially deep Learning (DL), to improve the detection of cyberattacks. DL-based IDSs require balanced datasets with large amounts of labeled data; however, there is a lack of such large collections in IoT networks. This paper proposes an efficient intrusion detection framework based on transfer learning (TL), knowledge transfer, and model refinement, for the effective detection of zero-day attacks. The framework is tailored to 5G IoT scenarios with unbalanced and scarce labeled datasets. The TL model is based on convolutional neural networks (CNNs). The framework was evaluated to detect a wide range of zero-day attacks. To this end, three specialized datasets were created. Experimental results show that the proposed TL-based framework achieves high accuracy and low false prediction rate (FPR). The proposed solution has better detection rates for the different families of known and zero-day attacks than any previous DL-based IDS. These results demonstrate that TL is effective in the detection of cyberattacks in IoT environments. [ABSTRACT FROM AUTHOR]

Cardiac arrhythmias pose a significant danger to human life; therefore, it is of utmost importance to be able to efficiently diagnose these arrhythmias promptly. There exist many techniques for the detection of arrhythmias; however, the most widely adopted method is the use of an Electrocardiogram (ECG). The manual analysis of ECGs by medical experts is often inefficient. Therefore, the detection and recognition of ECG characteristics via machine-learning techniques have become prevalent. There are two major drawbacks of existing machine-learning approaches: (a) they require extensive training time; and (b) they require manual feature selection. To address these issues, this paper presents a novel deep-learning framework that integrates various networks by stacking similar layers in each network to produce a single robust model. The proposed framework has been tested on two publicly available datasets for the recognition of five micro-classes of arrhythmias. The overall classification sensitivity, specificity, positive predictive value, and accuracy of the proposed approach are 98.37%, 99.59%, 98.41%, and 99.35%, respectively. The results are compared with state-of-the-art approaches. The proposed approach outperformed the existing approaches in terms of sensitivity, specificity, positive predictive value, accuracy and computational cost. [ABSTRACT FROM AUTHOR]

Nowadays in modern societies, a sedentary lifestyle is almost inevitable for a majority of the population. Long hours of sitting, especially in wrong postures, may result in health complications. A smart chair with the capability to identify sitting postures can help reduce health risks induced by a modern lifestyle. This paper presents the design, realization and evaluation of a new smart chair sensors system capable of sitting postures identification. The system consists of eight pressure sensors placed on the chair's sitting cushion and the backrest. A signal acquisition board was designed from scratch to acquire data generated by the pressure sensors and transmit them via a Wi-Fi network to a purposely developed graphical user interface which monitors and stores the acquired sensors' data on a computer. The designed system was tested by means of an extensive sitting experiment involving 40 subjects, and from the acquired data, the classification of the respective sitting postures out of eight possible postures was performed. Hereby, the performance of seven deep-learning algorithms was assessed. The best accuracy of 91.68% was achieved by an echo memory network model. The designed smart chair sensors system is simple and versatile, low cost and accurate, and it can easily be deployed in several smart chair environments, both for public and private contexts. [ABSTRACT FROM AUTHOR]

Recent advances in sensor technology are expected to lead to a greater use of wireless sensor networks (WSNs) in industry, logistics, healthcare, etc. On the other hand, advances in artificial intelligence (AI), machine learning (ML), and deep learning (DL) are becoming dominant solutions for processing large amounts of data from edge-synthesized heterogeneous sensors and drawing accurate conclusions with better understanding of the situation. Integration of the two areas WSN and AI has resulted in more accurate measurements, context-aware analysis and prediction useful for smart sensing applications. In this paper, a comprehensive overview of the latest developments in context-aware intelligent systems using sensor technology is provided. In addition, it also discusses the areas in which they are used, related challenges, motivations for adopting AI solutions, focusing on edge computing, i.e., sensor and AI techniques, along with analysis of existing research gaps. Another contribution of this study is the use of a semantic-aware approach to extract survey-relevant subjects. The latter specifically identifies eleven main research topics supported by the articles included in the work. These are analyzed from various angles to answer five main research questions. Finally, potential future research directions are also discussed. [ABSTRACT FROM AUTHOR]

In the oil and gas industries, predicting and classifying oil and gas production for hydrocarbon wells is difficult. Most oil and gas companies use reservoir simulation software to predict future oil and gas production and devise optimum field development plans. However, this process costs an immense number of resources and is time consuming. Each reservoir prediction experiment needs tens or hundreds of simulation runs, taking several hours or days to finish. In this paper, we attempt to overcome these issues by creating machine learning and deep learning models to expedite the process of forecasting oil and gas production. The dataset was provided by the leading oil producer, Saudi Aramco. Our approach reduced the time costs to a worst-case of a few minutes. Our study covered eight different ML and DL experiments and achieved its most outstanding R2 scores of 0.96 for XGBoost, 0.97 for ANN, and 0.98 for RNN over the other experiments. [ABSTRACT FROM AUTHOR]

The goal of this paper is to provide a Machine Learning-based solution that can be utilized to automate the Chicago Classification algorithm, the state-of-the-art scheme for esophageal motility disease identification. First, the photos were preprocessed by locating the area of interest—the precise instant of swallowing. After resizing and rescaling the photos, they were utilized as input for the Deep Learning models. The InceptionV3 Deep Learning model was used to identify the precise class of the IRP. We used the DenseNet201 CNN architecture to classify the images into 5 different classes of swallowing disorders. Finally, we combined the results of the two trained ML models to automate the Chicago Classification algorithm. With this solution we obtained a top-1 accuracy and f1-score of 86% with no human intervention, automating the whole flow, from image preprocessing until Chicago classification and diagnosis. [ABSTRACT FROM AUTHOR]

Artificial intelligence is serving as an impetus in digital health, clinical support, and health informatics for an informed patient's outcome. Previous studies only consider classification accuracies of cardiotocographic (CTG) datasets and disregard computational time, which is a relevant parameter in a clinical environment. This paper proposes a modified deep neural algorithm to classify untapped pathological and suspicious CTG recordings with the desired time complexity. In our newly developed classification algorithm, AlexNet architecture is merged with support vector machines (SVMs) at the fully connected layers to reduce time complexity. We used an open-source UCI (Machine Learning Repository) dataset of cardiotocographic (CTG) recordings. We divided 2126 CTG recordings into 3 classes (Normal, Pathological, and Suspected), including 23 attributes that were dynamically programmed and fed to our algorithm. We employed a deep transfer learning (TL) mechanism to transfer prelearned features to our model. To reduce time complexity, we implemented a strategy wherein layers in the convolutional base were partially trained to leave others in the frozen states. We used an ADAM optimizer for the optimization of hyperparameters. The presented algorithm also outperforms the leading architectures (RCNNs, ResNet, DenseNet, and GoogleNet) with respect to real-time accuracies, sensitivities, and specificities of 99.72%, 96.67%, and 99.6%, respectively, making it a viable candidate for clinical settings after real-time validation. [ABSTRACT FROM AUTHOR]

Heartbeat monitoring may play an essential role in the early detection of cardiovascular disease. When using a traditional monitoring system, an abnormal heartbeat may not appear during a recording in a healthcare facility due to the limited time. Thus, continuous and long-term monitoring is needed. Moreover, the conventional equipment may not be portable and cannot be used at arbitrary times and locations. A wearable sensor device such as Polar H10 offers the same capability as an alternative. It has gold-standard heartbeat recording and communication ability but still lacks analytical processing of the recorded data. An automatic heartbeat classification system can play as an analyzer and is still an open problem in the development stage. This paper proposes a heartbeat classifier based on RR interval data for real-time and continuous heartbeat monitoring using the Polar H10 wearable device. Several machine learning and deep learning methods were used to train the classifier. In the training process, we also compare intra-patient and inter-patient paradigms on the original and oversampling datasets to achieve higher classification accuracy and the fastest computation speed. As a result, with a constrain in RR interval data as the feature, the random forest-based classifier implemented in the system achieved up to 99.67% for accuracy, precision, recall, and F1-score. We are also conducting experiments involving healthy people to evaluate the classifier in a real-time monitoring system. [ABSTRACT FROM AUTHOR]

In this paper, we present a methodology based on generative adversarial network architecture to generate synthetic data sets with the intention of augmenting continuous glucose monitor data from individual patients. We use these synthetic data with the aim of improving the overall performance of prediction models based on machine learning techniques. Experiments were performed on two cohorts of patients suffering from type 1 diabetes mellitus with significant differences in their clinical outcomes. In the first contribution, we have demonstrated that the chosen methodology is able to replicate the intrinsic characteristics of individual patients following the statistical distributions of the original data. Next, a second contribution demonstrates the potential of synthetic data to improve the performance of machine learning approaches by testing and comparing different prediction models for the problem of predicting nocturnal hypoglycemic events in type 1 diabetic patients. The results obtained for both generative and predictive models are quite encouraging and set a precedent in the use of generative techniques to train new machine learning models. [ABSTRACT FROM AUTHOR]

This is a review focused on advances and current limitations of computer vision (CV) and how CV can help us obtain to more autonomous actions in surgery. It is a follow-up article to one that we previously published in Sensors entitled, "Artificial Intelligence Surgery: How Do We Get to Autonomous Actions in Surgery?" As opposed to that article that also discussed issues of machine learning, deep learning and natural language processing, this review will delve deeper into the field of CV. Additionally, non-visual forms of data that can aid computerized robots in the performance of more autonomous actions, such as instrument priors and audio haptics, will also be highlighted. Furthermore, the current existential crisis for surgeons, endoscopists and interventional radiologists regarding more autonomy during procedures will be discussed. In summary, this paper will discuss how to harness the power of CV to keep doctors who do interventions in the loop. [ABSTRACT FROM AUTHOR]

This paper proposes a novel inverse method based on the deep convolutional neural network (ConvNet) to extract snow's layer thickness and temperature via passive microwave remote sensing (PMRS). The proposed ConvNet is trained using simulated data obtained through conventional computational electromagnetic methods. Compared with the traditional inverse method, the trained ConvNet can predict the result with higher accuracy. Besides, the proposed method has a strong tolerance for noise. The proposed ConvNet composes three pairs of convolutional and activation layers with one additional fully connected layer to realize regression, i.e., the inversion of snow parameters. The feasibility of the proposed method in learning the inversion of snow parameters is validated by numerical examples. The inversion results indicate that the correlation coefficient ( R 2 ) ratio between the proposed ConvNet and conventional methods reaches 4.8, while the ratio for the root mean square error (RMSE) is only 0.18. Hence, the proposed method experiments with a novel path to improve the inversion of passive microwave remote sensing through deep learning approaches. [ABSTRACT FROM AUTHOR]

There is a group of users within the vehicular traffic ecosystem known as Vulnerable Road Users (VRUs). VRUs include pedestrians, cyclists, motorcyclists, among others. On the other hand, connected autonomous vehicles (CAVs) are a set of technologies that combines, on the one hand, communication technologies to stay always ubiquitous connected, and on the other hand, automated technologies to assist or replace the human driver during the driving process. Autonomous vehicles are being visualized as a viable alternative to solve road accidents providing a general safe environment for all the users on the road specifically to the most vulnerable. One of the problems facing autonomous vehicles is to generate mechanisms that facilitate their integration not only within the mobility environment, but also into the road society in a safe and efficient way. In this paper, we analyze and discuss how this integration can take place, reviewing the work that has been developed in recent years in each of the stages of the vehicle-human interaction, analyzing the challenges of vulnerable users and proposing solutions that contribute to solving these challenges. [ABSTRACT FROM AUTHOR]

A fetal ultrasound (US) is a technique to examine a baby's maturity and development. US examinations have varying purposes throughout pregnancy. Consequently, in the second and third trimester, US tests are performed for the assessment of Amniotic Fluid Volume (AFV), a key indicator of fetal health. Disorders resulting from abnormal AFV levels, commonly referred to as oligohydramnios or polyhydramnios, may pose a serious threat to a mother's or child's health. This paper attempts to accumulate and compare the most recent advancements in Artificial Intelligence (AI)-based techniques for the diagnosis and classification of AFV levels. Additionally, we provide a thorough and highly inclusive breakdown of other relevant factors that may cause abnormal AFV levels, including, but not limited to, abnormalities in the placenta, kidneys, or central nervous system, as well as other contributors, such as preterm birth or twin-to-twin transfusion syndrome. Furthermore, we bring forth a concise overview of all the Machine Learning (ML) and Deep Learning (DL) techniques, along with the datasets supplied by various researchers. This study also provides a brief rundown of the challenges and opportunities encountered in this field, along with prospective research directions and promising angles to further explore. [ABSTRACT FROM AUTHOR]

Traditional machine learning methods rely on the training data and target data having the same feature space and data distribution. The performance may be unacceptable if there is a difference in data distribution between the training and target data, which is called cross-domain learning problem. In recent years, many domain adaptation methods have been proposed to solve this kind of problems and make much progress. However, existing domain adaptation approaches have a common assumption that the number of the data in source domain (labeled data) and target domain (unlabeled data) is matched. In this paper, the scenarios in real manufacturing site are considered, that the target domain data is much less than source domain data at the beginning, but the number of target domain data will increase as time goes by. A novel method is proposed for fault diagnosis of rolling bearing with online imbalanced cross-domain data. Finally, the proposed method which is tested on bearing dataset (CWRU) has achieved prediction accuracy of 95.89% with only 40 target samples. The results have been compared with other traditional methods. The comparisons show that the proposed online domain adaptation fault diagnosis method has achieved significant improvements. In addition, the deep transfer learning model by adaptive- network-based fuzzy inference system (ANFIS) is introduced to interpretation the results. [ABSTRACT FROM AUTHOR]

Road segmentation has been one of the leading research areas in the realm of autonomous driving cars due to the possible benefits autonomous vehicles can offer. Significant reduction of crashes, greater independence for the people with disabilities, and reduced traffic congestion on the roads are some of the vivid examples of them. Considering the importance of self-driving cars, it is vital to develop models that can accurately segment drivable regions of roads. The recent advances in the area of deep learning have presented effective methods and techniques to tackle road segmentation tasks effectively. However, the results of most of them are not satisfactory for implementing them into practice. To tackle this issue, in this paper, we propose a novel model, dubbed as TA-Unet, that is able to produce quality drivable road region segmentation maps. The proposed model incorporates a triplet attention module into the encoding stage of the U-Net network to compute attention weights through the triplet branch structure. Additionally, to overcome the class-imbalance problem, we experiment on different loss functions, and confirm that using a mixed loss function leads to a boost in performance. To validate the performance and efficiency of the proposed method, we adopt the publicly available UAS dataset, and compare its results to the framework of the dataset and also to four state-of-the-art segmentation models. Extensive experiments demonstrate that the proposed TA-Unet outperforms baseline methods both in terms of pixel accuracy and mIoU, with 98.74% and 97.41%, respectively. Finally, the proposed method yields clearer segmentation maps on different sample sets compared to other baseline methods. [ABSTRACT FROM AUTHOR]

Diabetes mellitus (DM) is one of the most prevalent diseases in the world, and is correlated to a high index of mortality. One of its major complications is diabetic foot, leading to plantar ulcers, amputation, and death. Several studies report that a thermogram helps to detect changes in the plantar temperature of the foot, which may lead to a higher risk of ulceration. However, in diabetic patients, the distribution of plantar temperature does not follow a standard pattern, thereby making it difficult to quantify the changes. The abnormal temperature distribution in infrared (IR) foot thermogram images can be used for the early detection of diabetic foot before ulceration to avoid complications. There is no machine learning-based technique reported in the literature to classify these thermograms based on the severity of diabetic foot complications. This paper uses an available labeled diabetic thermogram dataset and uses the k-mean clustering technique to cluster the severity risk of diabetic foot ulcers using an unsupervised approach. Using the plantar foot temperature, the new clustered dataset is verified by expert medical doctors in terms of risk for the development of foot ulcers. The newly labeled dataset is then investigated in terms of robustness to be classified by any machine learning network. Classical machine learning algorithms with feature engineering and a convolutional neural network (CNN) with image-enhancement techniques are investigated to provide the best-performing network in classifying thermograms based on severity. It is found that the popular VGG 19 CNN model shows an accuracy, precision, sensitivity, F1-score, and specificity of 95.08%, 95.08%, 95.09%, 95.08%, and 97.2%, respectively, in the stratification of severity. A stacking classifier is proposed using extracted features of the thermogram, which is created using the trained gradient boost classifier, XGBoost classifier, and random forest classifier. This provides a comparable performance of 94.47%, 94.45%, 94.47%, 94.43%, and 93.25% for accuracy, precision, sensitivity, F1-score, and specificity, respectively. [ABSTRACT FROM AUTHOR]

Driving event detection and driver behavior recognition have been widely explored for many purposes, including detecting distractions, classifying driver actions, detecting kidnappings, pricing vehicle insurance, evaluating eco-driving, and managing shared and leased vehicles. Some systems can recognize the main driving events (e.g., accelerating, braking, and turning) by using in-vehicle devices, such as inertial measurement unit (IMU) sensors. In general, feature extraction is a commonly used technique to obtain robust and meaningful information from the sensor signals to guarantee the effectiveness of the subsequent classification algorithm. However, a general assessment of deep neural networks merits further investigation, particularly regarding end-to-end models based on Convolutional Neural Networks (CNNs), which combine two components, namely feature extraction and the classification parts. This paper primarily explores supervised deep-learning models based on 1D and 2D CNNs to classify driving events from the signals of linear acceleration and angular velocity obtained with the IMU sensors of a smartphone placed in the instrument panel of the vehicle. Aggressive and non-aggressive behaviors can be recognized by monitoring driving events, such as accelerating, braking, lane changing, and turning. The experimental results obtained are promising since the best classification model achieved accuracy values of up to 82.40%, and macro- and micro-average F1 scores, respectively, equal to 75.36% and 82.40%, thus, demonstrating high performance in the classification of driving events. [ABSTRACT FROM AUTHOR]

In recent times, organisations in a variety of businesses, such as healthcare, education, and others, have been using the Internet of Things (IoT) to produce more competent and improved services. The widespread use of IoT devices makes our lives easier. On the other hand, the IoT devices that we use suffer vulnerabilities that may impact our lives. These unsafe devices accelerate and ease cybersecurity attacks, specifically when using a botnet. Moreover, restrictions on IoT device resources, such as limitations in power consumption and the central processing unit and memory, intensify this issue because they limit the security techniques that can be used to protect IoT devices. Fortunately, botnets go through different stages before they can start attacks, and they can be detected in the early stage. This research paper proposes a framework focusing on detecting an IoT botnet in the early stage. An empirical experiment was conducted to investigate the behaviour of the early stage of the botnet, and then a baseline machine learning model was implemented for early detection. Furthermore, the authors developed an effective detection method, namely, Cross CNN_LSTM, to detect the IoT botnet based on using fusion deep learning models of a convolutional neural network (CNN) and long short-term memory (LSTM). According to the conducted experiments, the results show that the suggested model is accurate and outperforms some of the state-of-the-art methods, and it achieves 99.7 accuracy. Finally, the authors developed a kill chain model to prevent IoT botnet attacks in the early stage. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new privatization mechanism based on a naive theory of a perturbation on a probability using wavelets, such as a noise perturbs the signal of a digital image sensor. Wavelets are employed to extract information from a wide range of types of data, including audio signals and images often related to sensors, as unstructured data. Specifically, the cumulative wavelet integral function is defined to build the perturbation on a probability with the help of this function. We show that an arbitrary distribution function additively perturbed is still a distribution function, which can be seen as a privatized distribution, with the privatization mechanism being a wavelet function. Thus, we offer a mathematical method for choosing a suitable probability distribution for data by starting from some guessed initial distribution. Examples of the proposed method are discussed. Computational experiments were carried out using a database-sensor and two related algorithms. Several knowledge areas can benefit from the new approach proposed in this investigation. The areas of artificial intelligence, machine learning, and deep learning constantly need techniques for data fitting, whose areas are closely related to sensors. Therefore, we believe that the proposed privatization mechanism is an important contribution to increasing the spectrum of existing techniques. [ABSTRACT FROM AUTHOR]

Sentiment analysis was nominated as a hot research topic a decade ago for its increasing importance in analyzing the people's opinions extracted from social media platforms. Although the Arabic language has a significant share of the content shared across social media platforms, its content's sentiment analysis is still limited due to its complex morphological structures and the varieties of dialects. Traditional machine learning and deep neural algorithms have been used in a variety of studies to predict sentiment analysis. Therefore, a need of changing current mechanisms is required to increase the accuracy of sentiment analysis prediction. This paper proposed an optimized heterogeneous stacking ensemble model for enhancing the performance of Arabic sentiment analysis. The proposed model combines three different of pre-trained Deep Learning (DL) models: Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU) in conjunction with three meta-learners Logistic Regression (LR), Random Forest (RF), and Support Vector Machine (SVM) in order to enhance model's performance for predicting Arabic sentiment analysis. The performance of the proposed model with RNN, LSTM, GRU, and the five regular ML techniques: Decision Tree (DT), LR, K-Nearest Neighbor (KNN), RF, and Naive Bayes (NB) are compared using three benchmarks Arabic dataset. Parameters of Machine Learning (ML) and DL are optimized using Grid search and KerasTuner, respectively. Accuracy, precision, recall, and f1-score were applied to evaluate the performance of the models and validate the results. The results show that the proposed ensemble model has achieved the best performance for each dataset compared with other models. [ABSTRACT FROM AUTHOR]

In recent years, due to the advancements in machine learning, object detection has become a mainstream task in the computer vision domain. The first phase of object detection is to find the regions where objects can exist. With the improvements in deep learning, traditional approaches, such as sliding windows and manual feature selection techniques, have been replaced with deep learning techniques. However, object detection algorithms face a problem when performed in low light, challenging weather, and crowded scenes, similar to any other task. Such an environment is termed a challenging environment. This paper exploits pixel-level information to improve detection under challenging situations. To this end, we exploit the recently proposed hybrid task cascade network. This network works collaboratively with detection and segmentation heads at different cascade levels. We evaluate the proposed methods on three complex datasets of ExDark, CURE-TSD, and RESIDE, and achieve a mAP of 0.71, 0.52, and 0.43, respectively. Our experimental results assert the efficacy of the proposed approach. [ABSTRACT FROM AUTHOR]

Pixel-level image fusion is an effective way to fully exploit the rich texture information of visible images and the salient target characteristics of infrared images. With the development of deep learning technology in recent years, the image fusion algorithm based on this method has also achieved great success. However, owing to the lack of sufficient and reliable paired data and a nonexistent ideal fusion result as supervision, it is difficult to design a precise network training mode. Moreover, the manual fusion strategy has difficulty ensuring the full use of information, which easily causes redundancy and omittance. To solve the above problems, this paper proposes a multi-stage visible and infrared image fusion network based on an attention mechanism (MSFAM). Our method stabilizes the training process through multi-stage training and enhances features by the learning attention fusion block. To improve the network effect, we further design a Semantic Constraint module and Push–Pull loss function for the fusion task. Compared with several recently used methods, the qualitative comparison intuitively shows more beautiful and natural fusion results by our model with a stronger applicability. For quantitative experiments, MSFAM achieves the best results in three of the six frequently used metrics in fusion tasks, while other methods only obtain good scores on a single metric or a few metrics. Besides, a commonly used high-level semantic task, i.e., object detection, is used to prove its greater benefits for downstream tasks compared with single-light images and fusion results by existing methods. All these experiments prove the superiority and effectiveness of our algorithm. [ABSTRACT FROM AUTHOR]

Identifying accident patterns is one of the most vital research foci of driving analysis. Environmental or safety applications and the growing area of fleet management all benefit from accident detection contributions by minimizing the risk vehicles and drivers are subject to, improving their service and reducing overhead costs. Some solutions have been proposed in the past literature for automated accident detection that are mainly based on traffic data or external sensors. However, traffic data can be difficult to access, while external sensors can end up being difficult to set up and unreliable, depending on how they are used. Additionally, the scarcity of accident detection data has limited the type of approaches used in the past, leaving in particular, machine learning (ML) relatively unexplored. Thus, in this paper, we propose a ML framework for automated car accident detection based on mutimodal in-car sensors. Our work is a unique and innovative study on detecting real-world driving accidents by applying state-of-the-art feature extraction methods using basic sensors in cars. In total, five different feature extraction approaches, including techniques based on feature engineering and feature learning with deep learning are evaluated on the strategic highway research program (SHRP2) naturalistic driving study (NDS) crash data set. The main observations of this study are as follows: (1) CNN features with a SVM classifier obtain very promising results, outperforming all other tested approaches. (2) Feature engineering and feature learning approaches were finding different best performing features. Therefore, our fusion experiment indicates that these two feature sets can be efficiently combined. (3) Unsupervised feature extraction remarkably achieves a notable performance score. [ABSTRACT FROM AUTHOR]

The Internet of Things (IoT) is a widely used technology in automated network systems across the world. The impact of the IoT on different industries has occurred in recent years. Many IoT nodes collect, store, and process personal data, which is an ideal target for attackers. Several researchers have worked on this problem and have presented many intrusion detection systems (IDSs). The existing system has difficulties in improving performance and identifying subcategories of cyberattacks. This paper proposes a deep-convolutional-neural-network (DCNN)-based IDS. A DCNN consists of two convolutional layers and three fully connected dense layers. The proposed model aims to improve performance and reduce computational power. Experiments were conducted utilizing the IoTID20 dataset. The performance analysis of the proposed model was carried out with several metrics, such as accuracy, precision, recall, and F1-score. A number of optimization techniques were applied to the proposed model in which Adam, AdaMax, and Nadam performance was optimum. In addition, the proposed model was compared with various advanced deep learning (DL) and traditional machine learning (ML) techniques. All experimental analysis indicates that the accuracy of the proposed approach is high and more robust than existing DL-based algorithms. [ABSTRACT FROM AUTHOR]

The belt conveyor is an essential piece of equipment in coal mining for coal transportation, and its stable operation is key to efficient production. Belt surface of the conveyor is vulnerable to foreign bodies which can be extremely destructive. In the past decades, much research and numerous approaches to inspect belt status have been proposed, and machine learning-based non-destructive testing (NDT) methods are becoming more and more popular. Deep learning (DL), as a branch of machine learning (ML), has been widely applied in data mining, natural language processing, pattern recognition, image processing, etc. Generative adversarial networks (GAN) are one of the deep learning methods based on generative models and have been proved to be of great potential. In this paper, a novel multi-classification conditional CycleGAN (MCC-CycleGAN) method is proposed to generate and discriminate surface images of damages of conveyor belt. A novel architecture of improved CycleGAN is designed to enhance the classification performance using a limited capacity images dataset. Experimental results show that the proposed deep learning network can generate realistic belt surface images with defects and efficiently classify different damaged images of the conveyor belt surface. [ABSTRACT FROM AUTHOR]

The IETF Routing Over Low power and Lossy network (ROLL) working group defined IPv6 Routing Protocol for Low Power and Lossy Network (RPL) to facilitate efficient routing in IPv6 over Low-Power Wireless Personal Area Networks (6LoWPAN). Limited resources of 6LoWPAN nodes make it challenging to secure the environment, leaving it vulnerable to threats and security attacks. Machine Learning (ML) and Deep Learning (DL) approaches have shown promise as effective and efficient mechanisms for detecting anomalous behaviors in RPL-based 6LoWPAN. Therefore, this paper systematically reviews and critically analyzes the research landscape on ML, DL, and combined ML-DL approaches applied to detect attacks in RPL networks. In addition, this study examined existing datasets designed explicitly for the RPL network. This work collects relevant studies from five major databases: Google Scholar, Springer Link, Scopus, Science Direct, and IEEE Xplore® digital library. Furthermore, 15,543 studies, retrieved from January 2016 to mid-2021, were refined according to the assigned inclusion criteria and designed research questions resulting in 49 studies. Finally, a conclusive discussion highlights the issues and challenges in the existing studies and proposes several future research directions. [ABSTRACT FROM AUTHOR]

In aerospace, marine, and other heavy industries, bearing fault diagnosis has been an essential part of improving machine life, reducing economic losses, and avoiding safety problems caused by machine bearing failures. Most existing bearing fault diagnosis methods face challenges in extracting the fault features from raw bearing fault data. Compared with traditional methods for bearing fault characteristics extraction, deep neural networks can automatically extract intrinsic features without expert knowledge. The convolutional neural network (CNN) was utilized most widely in extracting representative features of bearing faults. Fundamental to this, the hybrid models based on the CNN and individual classifiers were proposed to diagnose bearing faults. However, CNN may not be suitable for all bearing fault classifiers. It is crucial to identify the classifiers which can maximize the CNN feature extraction ability. In this paper, four hybrid models based on CNN were built, and their fault detection accuracy and efficiency were compared. The comparative analysis showed that the random forest (RF) and support vector machine (SVM) could make full use of the CNN feature extraction ability. [ABSTRACT FROM AUTHOR]

Recently, the frequent occurrence of the misuse and intrusion of UAVs has made it a research challenge to identify and detect them effectively, and relatively high bandwidth and pressure on data transmission and real-time processing exist when sampling UAV communication signals using the RF detection method. In this paper, firstly, for data sampling, we chose a compressed sensing technique to replace the traditional sampling theorem and used a multi-channel random demodulator to sample the signal; secondly, for the detection and identification of the presence, type, and flight pattern of UAVs, a multi-stage deep learning-based UAV identification and detection method was proposed by exploiting the difference in communication signals between UAVs and controllers under different circumstances. The data samples are first passed by detectors that detect the presence of UAVs, then classifiers are used to identify the type of UAVs, and finally flight patterns are judged by the corresponding classifiers, for which two neural network structures (DNN and CNN) are constructed by deep learning algorithms and evaluated and validated by a 10-fold cross-validation method, with the DNN network used for detectors and the CNN network for subsequent type and flying mode classification. The experimental results demonstrate, first, the effectiveness of using compressed sensing for sampling the communication signals of UAVs and controllers; and second, the detecting method with multi-stage DL detects higher efficiency and accuracy compared with existing detecting methods, detecting the presence, type, and flight model of UAVs with an accuracy of over 99%. [ABSTRACT FROM AUTHOR]

Forward vehicle detection is the key technique to preventing car incidents in front. Artificial intelligence (AI) techniques are used to more accurately detect vehicles, but AI-based vehicle detection takes a lot of processing time due to its high computational complexity. When there is a risk of collision with a vehicle in front, the slow detection speed of the vehicle may lead to an accident. To quickly detect a vehicle in real-time, a high-speed and lightweight vehicle detection technique with similar detection performance to that of an existing AI-based vehicle detection is required. In addition, to apply forward collision warning system (FCWS) technology to vehicles, it is important to provide high performance based on low-power embedded systems because the vehicle's battery consumption must remain low. The vehicle detection algorithm occupies the most resources in FCWS. To reduce power consumption, it is important to reduce the computational complexity of an algorithm, that is, the amount of resources required to run it. This paper describes a method for fast, accurate forward vehicle detection using machine learning and deep learning. To detect a vehicle in consecutive images consistently, a Kalman filter is used to predict the bounding box based on the tracking algorithm and correct it based on the detection algorithm. As a result, its vehicle detection speed is about 25.85 times faster than deep-learning-based object detection is, and its detection accuracy is better than machine-learning-based object detection is. [ABSTRACT FROM AUTHOR]

Building contour extraction from high-resolution remote sensing images is a basic task for the reasonable planning of regional construction. Recently, building segmentation methods based on the U-Net network have become popular as they largely improve the segmentation accuracy by applying 'skip connection' to combine high-level and low-level feature information more effectively. Meanwhile, researchers have demonstrated that introducing an attention mechanism into U-Net can enhance local feature expression and improve the performance of building extraction in remote sensing images. In this paper, we intend to explore the effectiveness of the primeval attention gate module and propose the novel Attention Gate Module (AG) based on adjusting the position of 'Resampler' in an attention gate to Sigmoid function for a building extraction task, and a novel Attention Gates U network (AGs-Unet) is further proposed based on AG, which can automatically learn different forms of building structures in high-resolution remote sensing images and realize efficient extraction of building contour. AGs-Unet integrates attention gates with a single U-Net network, in which a series of attention gate modules are added into the 'skip connection' for suppressing the irrelevant and noisy feature responses in the input image to highlight the dominant features of the buildings in the image. AGs-Unet improves the feature selection of the attention map to enhance the ability of feature learning, as well as paying attention to the feature information of small-scale buildings. We conducted the experiments on the WHU building dataset and the INRIA Aerial Image Labeling dataset, in which the proposed AGs-Unet model is compared with several classic models (such as FCN8s, SegNet, U-Net, and DANet) and two state-of-the-art models (such as PISANet, and ARC-Net). The extraction accuracy of each model is evaluated by using three evaluation indexes, namely, overall accuracy, precision, and intersection over union. Experimental results show that the proposed AGs-Unet model can improve the quality of building extraction from high-resolution remote sensing images effectively in terms of prediction performance and result accuracy. [ABSTRACT FROM AUTHOR]

Distributed edge intelligence is a disruptive research area that enables the execution of machine learning and deep learning (ML/DL) algorithms close to where data are generated. Since edge devices are more limited and heterogeneous than typical cloud devices, many hindrances have to be overcome to fully extract the potential benefits of such an approach (such as data-in-motion analytics). In this paper, we investigate the challenges of running ML/DL on edge devices in a distributed way, paying special attention to how techniques are adapted or designed to execute on these restricted devices. The techniques under discussion pervade the processes of caching, training, inference, and offloading on edge devices. We also explore the benefits and drawbacks of these strategies. [ABSTRACT FROM AUTHOR]

Cardiovascular diseases are the leading cause of death globally, causing nearly 17.9 million deaths per year. Therefore, early detection and treatment are critical to help improve this situation. Many manufacturers have developed products to monitor patients' heart conditions as they perform their daily activities. However, very few can diagnose complex heart anomalies beyond detecting rhythm fluctuation. This paper proposes a new method that combines a Short-Time Fourier Transform (STFT) spectrogram of the ECG signal with handcrafted features to detect heart anomalies beyond commercial product capabilities. Using the proposed Convolutional Neural Network, the algorithm can detect 16 different rhythm anomalies with an accuracy of 99.79% with 0.15% false-alarm rate and 99.74% sensitivity. Additionally, the same algorithm can also detect 13 heartbeat anomalies with 99.18% accuracy with 0.45% false-alarm rate and 98.80% sensitivity. [ABSTRACT FROM AUTHOR]

A novel radar high-resolution range profile (HRRP) target recognition method based on a stacked autoencoder (SAE) and extreme learning machine (ELM) is presented in this paper. As a key component of deep structure, the SAE does not only learn features by making use of data, it also obtains feature expressions at different levels of data. However, with the deep structure, it is hard to achieve good generalization performance with a fast learning speed. ELM, as a new learning algorithm for single hidden layer feedforward neural networks (SLFNs), has attracted great interest from various fields for its fast learning speed and good generalization performance. However, ELM needs more hidden nodes than conventional tuning-based learning algorithms due to the random set of input weights and hidden biases. In addition, the existing ELM methods cannot utilize the class information of targets well. To solve this problem, a regularized ELM method based on the class information of the target is proposed. In this paper, SAE and the regularized ELM are combined to make full use of their advantages and make up for each of their shortcomings. The effectiveness of the proposed method is demonstrated by experiments with measured radar HRRP data. The experimental results show that the proposed method can achieve good performance in the two aspects of real-time and accuracy, especially when only a few training samples are available. [ABSTRACT FROM AUTHOR]

In this paper, we investigate various machine learning classifiers used in our Virtual Reality (VR) system for treating acrophobia. The system automatically estimates fear level based on multimodal sensory data and a self-reported emotion assessment. There are two modalities of expressing fear ratings: the 2-choice scale, where 0 represents relaxation and 1 stands for fear; and the 4-choice scale, with the following correspondence: 0—relaxation, 1—low fear, 2—medium fear and 3—high fear. A set of features was extracted from the sensory signals using various metrics that quantify brain (electroencephalogram—EEG) and physiological linear and non-linear dynamics (Heart Rate—HR and Galvanic Skin Response—GSR). The novelty consists in the automatic adaptation of exposure scenario according to the subject’s affective state. We acquired data from acrophobic subjects who had undergone an in vivo pre-therapy exposure session, followed by a Virtual Reality therapy and an in vivo evaluation procedure. Various machine and deep learning classifiers were implemented and tested, with and without feature selection, in both a user-dependent and user-independent fashion. The results showed a very high cross-validation accuracy on the training set and good test accuracies, ranging from 42.5% to 89.5%. The most important features of fear level classification were GSR, HR and the values of the EEG in the beta frequency range. For determining the next exposure scenario, a dominant role was played by the target fear level, a parameter computed by taking into account the patient’s estimated fear level. [ABSTRACT FROM AUTHOR]

Across the globe, remote image data is rapidly being collected for the assessment of benthic communities from shallow to extremely deep waters on continental slopes to the abyssal seas. Exploiting this data is presently limited by the time it takes for experts to identify organisms found in these images. With this limitation in mind, a large effort has been made globally to introduce automation and machine learning algorithms to accelerate both classification and assessment of marine benthic biota. One major issue lies with organisms that move with swell and currents, such as kelps. This paper presents an automatic hierarchical classification method local binary classification as opposed to the conventional flat classification to classify kelps in images collected by autonomous underwater vehicles. The proposed kelp classification approach exploits learned feature representations extracted from deep residual networks. We show that these generic features outperform the traditional off-the-shelf CNN features and the conventional hand-crafted features. Experiments also demonstrate that the hierarchical classification method outperforms the traditional parallel multi-class classifications by a significant margin (90.0% vs. 57.6% and 77.2% vs. 59.0%) on Benthoz15 and Rottnest datasets respectively. Furthermore, we compare different hierarchical classification approaches and experimentally show that the sibling hierarchical training approach outperforms the inclusive hierarchical approach by a significant margin. We also report an application of our proposed method to study the change in kelp cover over time for annually repeated AUV surveys. [ABSTRACT FROM AUTHOR]

Adopting deep learning methods for human activity recognition has been effective in extracting discriminative features from raw input sequences acquired from body-worn sensors. Although human movements are encoded in a sequence of successive samples in time, typical machine learning methods perform recognition tasks without exploiting the temporal correlations between input data samples. Convolutional neural networks (CNNs) address this issue by using convolutions across a one-dimensional temporal sequence to capture dependencies among input data. However, the size of convolutional kernels restricts the captured range of dependencies between data samples. As a result, typical models are unadaptable to a wide range of activity-recognition configurations and require fixed-length input windows. In this paper, we propose the use of deep recurrent neural networks (DRNNs) for building recognition models that are capable of capturing long-range dependencies in variable-length input sequences. We present unidirectional, bidirectional, and cascaded architectures based on long short-term memory (LSTM) DRNNs and evaluate their effectiveness on miscellaneous benchmark datasets. Experimental results show that our proposed models outperform methods employing conventional machine learning, such as support vector machine (SVM) and k-nearest neighbors (KNN). Additionally, the proposed models yield better performance than other deep learning techniques, such as deep believe networks (DBNs) and CNNs. [ABSTRACT FROM AUTHOR]

In this paper, I propose a bird eye view image detection method for parking areas and collision risk areas at the same time in parking situations. Deep learning algorithms using area detection and semantic segmentation were used. The main architecture of the method described in this paper is based on a harmonic densely connected network and a cross-stage partial network. The dataset used for training was calibrated to four 190° wide-angle cameras to generate around view monitor (AVM) images based on the Chungbuk National University parking lot, and an experiment was performed based on this dataset. In the experimental results, the available parking area was visualized by detecting the parking line, parking area, and available driving area in the AVM images. Furthermore, the undetected area in the semantic segmentation as a collision risk area was visualized in order to obtain the results. According to the proposed attention CSPHarDNet model, the experimental results were 81.89% mIoU and 18.36 FPS in a NVIDIA Xavier environment. The results of this experiment demonstrated that algorithms can be used in real time in a parking situation and have better performance results compared to the conventional HarDNet. [ABSTRACT FROM AUTHOR]

Bearings are widely used in various electrical and mechanical equipment. As their core components, failures often have serious consequences. At present, most parameter adjustment methods are still manual adjustments of parameters. This adjustment method is easily affected by prior knowledge, easily falls into the local optimal solution, cannot obtain the global optimal solution, and requires a lot of resources. Therefore, this paper proposes a new method for bearing fault diagnosis based on wavelet packet transform and convolutional neural network optimized by a simulated annealing algorithm. Firstly, the original bearing vibration signal is extracted by wavelet packet transform to obtain the spectrogram, and then the obtained spectrogram is sent to the convolutional neural network for parameter adjustment, and finally the simulated annealing algorithm is used to adjust the parameters. To verify the effectiveness of the method, the bearing database of Case Western Reserve University is used for testing, and the traditional intelligent bearing fault diagnosis methods are compared. The results show that the new method for bearing fault diagnosis proposed in this paper has a better and more reliable diagnosis effect than the existing machine learning and deep learning methods. [ABSTRACT FROM AUTHOR]

Neurodevelopmental disorders (NDD) are impairments of the growth and development of the brain and/or central nervous system. In the light of clinical findings on early diagnosis of NDD and prompted by recent advances in hardware and software technologies, several researchers tried to introduce automatic systems to analyse the baby's movement, even in cribs. Traditional technologies for automatic baby motion analysis leverage contact sensors. Alternatively, remotely acquired video data (e.g., RGB or depth) can be used, with or without active/passive markers positioned on the body. Markerless approaches are easier to set up and maintain (without any human intervention) and they work well on non-collaborative users, making them the most suitable technologies for clinical applications involving children. On the other hand, they require complex computational strategies for extracting knowledge from data, and then, they strongly depend on advances in computer vision and machine learning, which are among the most expanding areas of research. As a consequence, also markerless video-based analysis of movements in children for NDD has been rapidly expanding but, to the best of our knowledge, there is not yet a survey paper providing a broad overview of how recent scientific developments impacted it. This paper tries to fill this gap and it lists specifically designed data acquisition tools and publicly available datasets as well. Besides, it gives a glimpse of the most promising techniques in computer vision, machine learning and pattern recognition which could be profitably exploited for children motion analysis in videos. [ABSTRACT FROM AUTHOR]