Search

In the traditional text detection process, the text area of the small receptive field in the video image is easily ignored, the features that can be extracted are few, and the calculation is large. These problems are not conducive to the recognition of text information. In this paper, a lightweight network structure on the basis of the EAST algorithm, the Convolution Block Attention Module (CBAM), is proposed. It is suitable for the spatial and channel hybrid attention module of text feature extraction of the natural scene video images. The improved structure proposed in this paper can obtain deep network features of text and reduce the computation of text feature extraction. Additionally, a hybrid feature pyramid + BLSTM network is designed to improve the attention to the small acceptance domain text regions and the text sequence features of the region. The test results on the ICDAR2015 demonstrate that the improved construction can effectively boost the attention of small acceptance domain text regions and improve the sequence feature detection accuracy of small acceptance domain of long text regions without significantly increasing computation. At the same time, the proposed network constructions are superior to the traditional EAST algorithm and other improved algorithms in accuracy rate P, recall rate R, and F-value. [ABSTRACT FROM AUTHOR]

Indoor localization has long been a challenging task due to the complexity and dynamism of indoor environments. This paper proposes ULG-SLAM, a novel unsupervised learning and geometric-based visual SLAM algorithm for robot localizability estimation to improve the accuracy and robustness of visual SLAM. Firstly, a dynamic feature filtering based on unsupervised learning and moving consistency checks is developed to eliminate the features of dynamic objects. Secondly, an improved line feature extraction algorithm based on LSD is proposed to optimize the effect of geometric feature extraction. Thirdly, geometric features are used to optimize localizability estimation, and an adaptive weight model and attention mechanism are built using the method of region delimitation and region growth. Finally, to verify the effectiveness and robustness of localizability estimation, multiple indoor experiments using the EuRoC dataset and TUM RGB-D dataset are conducted. Compared with ORBSLAM2, the experimental results demonstrate that absolute trajectory accuracy can be improved by 95% for equivalent processing speed in walking sequences. In fr3/walking_xyz and fr3/walking_half, ULG-SLAM tracks more trajectories than DS-SLAM, and the ATE RMSE is improved by 36% and 6%, respectively. Furthermore, the improvement in robot localizability over DynaSLAM is noteworthy, coming in at about 11% and 3%, respectively. [ABSTRACT FROM AUTHOR]

Extracting effective features from high-dimensional datasets is crucial for determining the accuracy of regression and classification models. Model predictions based on causality are known for their robustness. Thus, this paper introduces causality into feature selection and utilizes Feature Selection based on NOTEARS causal discovery (FSNT) for effective feature extraction. This method transforms the structural learning algorithm into a numerical optimization problem, enabling the rapid identification of the globally optimal causality diagram between features and the target variable. To assess the effectiveness of the FSNT algorithm, this paper evaluates its performance by employing 10 regression algorithms and 8 classification algorithms for regression and classification predictions on six real datasets from diverse fields. These results are then compared with three mainstream feature selection algorithms. The results indicate a significant average decline of 54.02% in regression prediction achieved by the FSNT algorithm. Furthermore, the algorithm exhibits exceptional performance in classification prediction, leading to an enhancement in the precision value. These findings highlight the effectiveness of FSNT in eliminating redundant features and significantly improving the accuracy of model predictions. [ABSTRACT FROM AUTHOR]

With the gradual popularization of autonomous driving technology, how to obtain traffic sign information efficiently and accurately is very important for subsequent decision-making and planning tasks. Traffic sign detection and recognition (TSDR) algorithms include color-based, shape-based, and machine learning based. However, the algorithms mentioned above are insufficient for traffic sign detection tasks in complex environments. In this paper, we propose a traffic sign detection and recognition paradigm based on deep learning algorithms. First, to solve the problem of insufficient spatial information in high-level features of small traffic signs, the parallel deformable convolution module (PDCM) is proposed in this paper. PDCM adaptively acquires the corresponding receptive field preserving the integrity of the abstract information through symmetrical branches thereby improving the feature extraction capability. Simultaneously, we propose sub-pixel convolution attention module (SCAM) based on the attention mechanism to alleviate the influence of scale distribution. Distinguishing itself from other feature fusion, our proposed method can better focus on the information of scale distribution through the attention module. Eventually, we introduce GSConv to further reduce the computational complexity of our proposed algorithm, better satisfying industrial application. Experimental results demonstrate that our proposed methods can effectively improve performance, both in detection accuracy and mAP@0.5. Specifically, when the proposed PDCM, SCAM, and GSConv are applied to the Yolov5, it achieves 89.2% mAP@0.5 in TT100K, which exceeds the benchmark network by 4.9%. [ABSTRACT FROM AUTHOR]

With an increase in the use of software, the incidence of bugs and resulting maintenance costs also increase. In open source projects, developer reassignment accounts for approximately 50%. Software maintenance costs can be reduced if appropriate developers are recommended to resolve bugs. In this study, features are extracted by applying feature selection for each developer. These features are entered into CNN-LSTM algorithm to learn the model and recommend appropriate developers. To compare the performance of the proposed model, open source projects (Google Chrome, Mozilla Core, and Mozilla Firefox) were used to compare the performance of the proposed method with a baseline for developer recommendation. In this paper, the performance showed 54% for F-measure and 52% for accuracy in open source projects. The proposed model has improved and showed about a 13% more effective performance improvement than with DeepTriage. It was discovered that the performance of the proposed model was better. [ABSTRACT FROM AUTHOR]

In actual driving scenes, recognizing and preventing drivers' non-standard driving behavior is helpful in reducing traffic accidents. To resolve the problems of various driving behaviors, a large range of action, and the low recognition accuracy of traditional detection methods, in this paper, a driving behavior recognition algorithm was proposed that combines an attention mechanism and lightweight network. The attention module was integrated into the YOLOV4 model after improving the feature extraction network, and the structure of the attention module was also improved. According to the 20,000 images of the Kaggle dataset, 10 typical driving behaviors were analyzed, processed, and recognized. The comparison and ablation experimental results showed that the fusion of an improved attention mechanism and lightweight network model had good performance in accuracy, model size, and FLOPs. [ABSTRACT FROM AUTHOR]

Internet-enabled (IoT) devices are typically small, low-powered devices used for sensing and computing that enable remote monitoring and control of various environments through the Internet. Despite their usefulness in achieving a more connected cyber-physical world, these devices are vulnerable to ransomware attacks due to their limited resources and connectivity. To combat these threats, machine learning (ML) can be leveraged to identify and prevent ransomware attacks on IoT devices before they can cause significant damage. In this research paper, we explore the use of ML techniques to enhance ransomware defense in IoT devices running on the PureOS operating system. We have developed a ransomware detection framework using machine learning, which combines the XGBoost and ElasticNet algorithms in a hybrid approach. The design and implementation of our framework are based on the evaluation of various existing machine learning techniques. Our approach was tested using a dataset of real-world ransomware attacks on IoT devices and achieved high accuracy (90%) and low false-positive rates, demonstrating its effectiveness in detecting and preventing ransomware attacks on IoT devices running PureOS. [ABSTRACT FROM AUTHOR]

Structural health monitoring (SHM) systems are used to analyze the health of infrastructures such as bridges, using data from various types of sensors. While SHM systems consist of various stages, feature extraction and pattern recognition steps are the most important. Consequently, signal processing techniques in the feature extraction stage and machine learning algorithms in the pattern recognition stage play an effective role in analyzing the health of bridges. In other words, there exists a plethora of signal processing techniques and machine learning algorithms, and the selection of the appropriate technique/algorithm is guided by the limitations of each technique/algorithm. The selection also depends on the requirements of SHM in terms of damage identification level and operating conditions. This has provided the motivation to conduct a Systematic literature review (SLR) of feature extraction techniques and pattern recognition algorithms for the structural health monitoring of bridges. The existing literature reviews describe the current trends in the field with different focus aspects. However, a systematic literature review that presents an in-depth comparative study of different applications of machine learning algorithms in the field of SHM of bridges does not exist. Furthermore, there is a lack of analytical studies that investigate the SHM systems in terms of several design considerations including feature extraction techniques, analytical approaches (classification/ regression), operational functionality levels (diagnosis/prognosis) and system implementation techniques (data-driven/model-based). Consequently, this paper identifies 45 recent research practices (during 2016–2023), pertaining to feature extraction techniques and pattern recognition algorithms in SHM for bridges through an SLR process. First, the identified research studies are classified into three different categories: supervised learning algorithms, neural networks and a combination of both. Subsequently, an in-depth analysis of various machine learning algorithms is performed in each category. Moreover, the analysis of selected research studies (total = 45) in terms of feature extraction techniques is made, and 25 different techniques are identified. Furthermore, this article also explores other design considerations like analytical approaches in the pattern recognition process, operational functionality and system implementation. It is expected that the outcomes of this research may facilitate the researchers and practitioners of the domain during the selection of appropriate feature extraction techniques, machine learning algorithms and other design considerations according to the SHM system requirements. [ABSTRACT FROM AUTHOR]

The failure of insulators may seriously threaten the safe operation of the power system, where the state detection of high-voltage insulators is a must for the normal and safe operation of the power system. Based on the data of insulators in aerial images, this work explored an enhanced particle swarm algorithm to optimize the parameters of the support vector machine. A support vector machine model was therefore established for the identification of the normal and defective states of the insulators. This methodology works with the structure minimization principle of SVM and the characteristics of particle swarm fast optimization. First, the aerial insulator image was segmented as a target by way of the seed region growth based on double-layer cascade morphological improvements, and then, HOG features plus GLCM features were extracted as sample data. Finally, an ameliorated PSO-SVM classifier was designed to realize insulator state identification. Comparisons were made between PSO-SVM and conventional machine learning algorithms, SVM and Random Forest, and an optimization algorithm, Gray Wolf Optimization Support Vector Machine (GWO-SVM), and advanced neural network CNN. The experimental results showed that the performance of the algorithm proposed in this paper touched the top level, where the recognition accuracy rate was 92.11%, the precision rate 90%, the recall rate 94.74%, and the F1-score 92.31%. [ABSTRACT FROM AUTHOR]

Indoor 3D positioning is useful in multistory buildings, such as shopping malls, libraries, and airports. This study focuses on indoor 3D positioning using wireless access points (AP) in an environment without adding additional hardware facilities in large-scale complex places. The integration of a deep learning algorithm into indoor 3D positioning is studied, and a 3D dynamic positioning model based on temporal fingerprints is proposed. In contrast to the traditional positioning models with a single input, the proposed method uses a sliding time window to build a temporal fingerprint chip as the input of the positioning model to provide abundant information for positioning. Temporal information can be used to distinguish locations with similar fingerprint vectors and to improve the accuracy and robustness of positioning. Moreover, deep learning has been applied for the automatic extraction of spatiotemporal features. A temporal convolutional network (TCN) feature extractor is proposed in this paper, which adopts a causal convolution mechanism, dilated convolution mechanism, and residual connection mechanism and is not limited by the size of the convolution kernel. It is capable of learning hidden information and spatiotemporal relationships from the input features and the extracted spatiotemporal features are connected with a deep neural network (DNN) regressor to fit the complex nonlinear mapping relationship between the features and position coordinates to estimate the 3D position coordinates of the target. Finally, an open-source public dataset was used to verify the performance of the localization algorithm. Experimental results demonstrated the effectiveness of the proposed positioning model and a comparison between the proposed model and existing models proved that the proposed model can provide more accurate three-dimensional position coordinates. [ABSTRACT FROM AUTHOR]

The scenario is very important to smartphone-based pedestrian positioning services. The smartphone is equipped with MEMS(Micro Electro Mechanical System) sensors, which have low accuracy. Now, the methods for scenario recognition are mainly machine-learning methods. The recognition rate of a single method is not high. Multi-model fusion can improve recognition accuracy, but it needs to collect many samples, the computational cost is high, and it is heavily dependent on feature selection. Therefore, we designed the DT-BP(decision tree-Bayesian probability) scenario recognition algorithm by introducing the Bayesian state transition model based on experience design in the decision tree. The decision-tree rules and state transition probability assignment methods were respectively designed for smartphone mode and motion mode. We carried out experiments for each scenario and compared them with the methods in the references. The results showed that the method proposed in this paper has a high recognition accuracy, which is equivalent to the accuracy of multi-model machine learning, but it is simpler, easier to implement, requires less computation, and requires fewer samples. [ABSTRACT FROM AUTHOR]

To solve the problems of low accuracy, low real-time performance, poor robustness and others caused by the complex environment, this paper proposes a face mask recognition and standard wear detection algorithm based on the improved YOLO-v4. Firstly, an improved CSPDarkNet53 is introduced into the trunk feature extraction network, which reduces the computing cost of the network and improves the learning ability of the model. Secondly, the adaptive image scaling algorithm can reduce computation and redundancy effectively. Thirdly, the improved PANet structure is introduced so that the network has more semantic information in the feature layer. At last, a face mask detection data set is made according to the standard wearing of masks. Based on the object detection algorithm of deep learning, a variety of evaluation indexes are compared to evaluate the effectiveness of the model. The results of the comparations show that the mAP of face mask recognition can reach 98.3% and the frame rate is high at 54.57 FPS, which are more accurate compared with the exiting algorithm. [ABSTRACT FROM AUTHOR]

Eye fatigue has a fatiguing effect on the eye muscles, and eye movement performance is a macroscopic response to the eye fatigue state. To detect and prevent the risk of eye fatigue in advance, this study designed an eye fatigue detection experiment, collected experimental data samples, and constructed experimental data sets. In this study, eye-tracking feature extraction was completed, and the significance difference of eye-tracking features under different fatigue states was discussed by two-way repeated-measures ANOVA (Analysis of Variance). The experimental results demonstrate the feasibility of eye fatigue detection from eye-tracking signals. In addition, this study considers the effects of different feature extraction methods on eye fatigue detection accuracy. This study examines the performance of machine learning algorithms based on manual feature calculation (SVM, DT, RM, ET) and deep learning algorithms based on automatic feature extraction (CNN, auto-encoder, transformer) in eye fatigue detection. Based on the combination of the methods, this study proposes the feature union auto-encoder algorithm, and the accuracy of the algorithm for eye fatigue detection on the experimental dataset is improved from 82.4% to 87.9%. [ABSTRACT FROM AUTHOR]

Nowadays, many image processing and machine learning (ML) methods are used in mango-quality classification systems. Existing algorithms perform independently to capture the relationships between features in the dataset. Nevertheless, each method has its advantages and disadvantages. As a result, this study presents an ensemble-learning framework that combines the strengths of multiple ML algorithms to make predictions. Initially, different image processing algorithms are used to extract external mango features. Next, the dataset is constructed by combining those features with weight values from sensor signals. Following that, different ML algorithms are evaluated on the dataset to determine which ones are robust. Subsequently, various ensemble-learning approaches are deployed, such as bagging, boosting, and stacking. Finally, those models are evaluated and compared, to decide which model is suited for this study's dataset. In the experimental part, the assessment of errors demonstrates the usefulness of image processing algorithms. Furthermore, evaluation of the training models revealed that the stacking model, which integrates several methods in both the base learner and meta-learner, produced the highest results in precision, recall, F1-score, and accuracy, with values of 0.9855, 0.9901, 0.9876, and 0.9863, respectively. These experimental results confirm the robustness of the proposed methodologies for classifying mango quality. [ABSTRACT FROM AUTHOR]

We present an algorithm to find corresponding authors of patents and scientific articles. The authors are given as records in Scopus and the Chinese Patents Database. This issue is known as the record linkage problem, defined as finding and linking individual records from separate databases that refer to the same real-world entity. The presented solution is based on a record linkage framework combined with text feature extraction and machine learning techniques. The main challenges were low data quality, lack of common record identifiers, and a limited number of other attributes shared by both data sources. Matching based solely on an exact comparison of authors' names does not solve the records linking problem because many Chinese authors share the same full name. Moreover, the English spelling of Chinese names is not standardized in the analyzed data. Three ideas on how to extend attribute sets and improve record linkage quality were proposed: (1) fuzzy matching of names, (2) comparison of abstracts of patents and articles, (3) comparison of scientists' main research areas calculated using all metadata available. The presented solution was evaluated in terms of matching quality and complexity on ≈250,000 record pairs linked by human experts. The results of numerical experiments show that the proposed strategies increase the quality of record linkage compared to typical solutions. [ABSTRACT FROM AUTHOR]

The deep learning feature extraction method and extreme learning machine (ELM) classification method are combined to establish a depth extreme learning machine model for wood image defect detection. The convolution neural network (CNN) algorithm alone tends to provide inaccurate defect locations, incomplete defect contour and boundary information, and inaccurate recognition of defect types. The nonsubsampled shearlet transform (NSST) is used here to preprocess the wood images, which reduces the complexity and computation of the image processing. CNN is then applied to manage the deep algorithm design of the wood images. The simple linear iterative clustering algorithm is used to improve the initial model; the obtained image features are used as ELM classification inputs. ELM has faster training speed and stronger generalization ability than other similar neural networks, but the random selection of input weights and thresholds degrades the classification accuracy. A genetic algorithm is used here to optimize the initial parameters of the ELM to stabilize the network classification performance. The depth extreme learning machine can extract high-level abstract information from the data, does not require iterative adjustment of the network weights, has high calculation efficiency, and allows CNN to effectively extract the wood defect contour. The distributed input data feature is automatically expressed in layer form by deep learning pre-training. The wood defect recognition accuracy reached 96.72% in a test time of only 187 ms. [ABSTRACT FROM AUTHOR]

Non-cooperative space object pose estimation is a key technique for spatial on-orbit servicing, where pose estimation algorithms based on low-quality, low-power monocular sensors provide a practical solution for spaceborne applications. The current pose estimation methods for non-cooperative space objects using monocular vision generally consist of three stages: object detection, landmark regression, and perspective-n-point (PnP) solver. However, there are drawbacks, such as low detection efficiency and the need for prior knowledge. To solve the above problems, an end-to-end non-cooperative space object pose estimation learning algorithm based on dual-channel transformer is proposed, a feature extraction backbone network based on EfficientNet is established, and two pose estimation subnetworks based on transformer are also established. A quaternion SoftMax-like activation function is designed to improve the precision of orientation error estimating. The method only uses RGB images, eliminating the need for a CAD model of the satellite, and simplifying the detection process by using an end-to-end network to directly detect satellite pose information. Experiments are carried out on the SPEED dataset provided by the European Space Agency (ESA). The results show that the proposed algorithm can successfully predict the satellite pose information and effectively decouple the spatial translation information and orientation information, which significantly improves the recognition efficiency compared with other methods. [ABSTRACT FROM AUTHOR]

Scientific analyses of urban flood risks are essential for evaluating urban flood insurance and designing drainage projects. Although the current rainfall monitoring system in China has a dense station network and high-precision rainfall data, the time series is short. In contrast, historical rainfall data have a longer sample time series but lower precision. This study introduced a PCA algorithm to reconstruct historical rainfall data. Based on the temporal and spatial characteristics of rainfall extracted from high-resolution rainfall data over the past decade, historical (6 h intervals) rainfall spatial data were reconstructed into high-resolution (1 h intervals) spatial data to satisfy the requirements of the urban flood risk analysis. The results showed that the average error between the reconstructed data and measured values in the high-value area was within 15% and in the low-value area was within 20%, representing decreases of approximately 65% and 40%, respectively, compared to traditional interpolation data. The reconstructed historical spatial rainfall data conformed to the temporal and spatial distribution characteristics of rainfall, improved the granularity of rainfall spatial data, and enabled the effective and reasonable extraction and summary of the fine temporal and spatial distribution characteristics of rainfall. [ABSTRACT FROM AUTHOR]

This research proposes a new feature extraction algorithm using aggregated user engagements on social media in order to achieve demographics and personality discovery tasks. Our proposed framework can discover seven essential attributes, including gender identity, age group, residential area, education level, political affiliation, religious belief, and personality type. Multiple feature sets are developed, including comment text, community activity, and hybrid features. Various machine learning algorithms are explored, such as support vector machines, random forest, multi-layer perceptron, and naïve Bayes. An empirical analysis is performed on various aspects, including correctness, robustness, training time, and the class imbalance problem. We obtained the highest prediction performance by using our proposed feature extraction algorithm. The result on personality type prediction was 87.18%. For the demographic attribute prediction task, our feature sets also outperformed the baseline at 98.1% for residential area, 94.7% for education level, 92.1% for gender identity, 91.5% for political affiliation, 60.6% for religious belief, and 52.0% for the age group. Moreover, this paper provides the guideline for the choice of classifiers with appropriate feature sets. [ABSTRACT FROM AUTHOR]

In this paper, a Deep Learning approach is proposed to classify impact data based on the type of impact (Hard or Soft Impacts), via obtaining voltage signals from Piezo-Electric sensors, mounted on a composite panel. The data is processed further to be classified based on their energy, location and material. Minimalistic and Automated feature extraction and selection is achieved via a deep learning algorithm. Convolutional Neural Networks (CNN) are employed to extract and select important features from the voltage data. Once features are selected the impacts, are classified based on either, Hard Impacts (simulated from steel impactors in a lab setting), Soft Impacts (simulated from silicon impactors in a lab setting) and their corresponding location and energy levels. Furthermore, in order to use the right data for training they are obtained from the signals as anomalies via Isolation Forests (IF) to speed up the process. Using this approach Hard and Soft Impacts, their corresponding locations and respective energies are identified with high accuracy. [ABSTRACT FROM AUTHOR]

Bearing spall detection and predicting its size are great challenges. Model-based simulation is a well-known traditional approach to physically model the influence of the spall on the bearing. Building a physical model is challenging due to the bearing complexity and the expert knowledge required to build such a model. Obviously, building a partial physical model for some of the spall sizes is easier. In this paper, we propose a machine-learning algorithm, called Probability-Based Forest, that uses a partial physical model. First, the behavior of some of the spall sizes is physically modeled and a simulator based on this model generates scenarios for these spall sizes in different conditions. Then, the machine-learning algorithm trains these scenarios to generate a prediction model of spall sizes even for those that have not been modeled by the physical model. Feature extraction is a key factor in the success of this approach. We extract features using two traditional approaches: statistical and physical, and an additional new approach: Time Series FeatuRe Extraction based on Scalable Hypothesis tests (TSFRESH). Experimental evaluation with well-known physical model shows that our approach achieves high accuracy, even in cases that have not been modeled by the physical model. Also, we show that the TSFRESH feature-extraction approach achieves the highest accuracy. [ABSTRACT FROM AUTHOR]

Orthogonal transformations, proper decomposition, and the Moore–Penrose inverse are traditional methods of obtaining the output layer weights for an extreme learning machine autoencoder. However, an increase in the number of hidden neurons causes higher convergence times and computational complexity, whereas the generalization capability is low when the number of neurons is small. One way to address this issue is to use the fast iterative shrinkage-thresholding algorithm (FISTA) to minimize the output weights of the extreme learning machine. In this work, we aim to improve the convergence speed of FISTA by using two fast algorithms of the shrinkage-thresholding class, called greedy FISTA (G-FISTA) and linearly convergent FISTA (LC-FISTA). Our method is an exciting proposal for decision-making involving the resolution of many application problems, especially those requiring longer computational times. In our experiments, we adopt six public datasets that are frequently used in machine learning: MNIST, NORB, CIFAR10, UMist, Caltech256, and Stanford Cars. We apply several metrics to evaluate the performance of our method, and the object of comparison is the FISTA algorithm due to its popularity for neural network training. The experimental results show that G-FISTA and LC-FISTA achieve higher convergence speeds in the autoencoder training process; for example, in the Stanford Cars dataset, G-FISTA and LC-FISTA are faster than FISTA by 48.42% and 47.32%, respectively. Overall, all three algorithms maintain good values of the performance metrics on all databases. [ABSTRACT FROM AUTHOR]

The air-door is an important device for adjusting the air flow in a mine. It opens and closes within a short time owing to transportation and other factors. Although the switching sensor alone can identify the air-door opening and closing, it cannot relate it to abnormal fluctuations in the wind speed. Large fluctuations in the wind-velocity sensor data during this time can lead to false alarms. To overcome this problem, we propose a method for identifying air-door opening and closing using a single wind-velocity sensor. A multi-scale sliding window (MSSW) is employed to divide the samples. Then, the data global features and fluctuation features are extracted using statistics and the discrete wavelet transform (DWT). In addition, a machine learning model is adopted to classify each sample. Further, the identification results are selected by merging the classification results using the non-maximum suppression method. Finally, considering the safety accidents caused by the air-door opening and closing in an actual production mine, a large number of experiments were carried out to verify the effect of the algorithm using a simulated tunnel model. The results show that the proposed algorithm exhibits superior performance when the gradient boosting decision tree (GBDT) is selected for classification. In the data set composed of air-door opening and closing experimental data, the accuracy, precision, and recall rates of the air-door opening and closing identification are 91.89%, 93.07%, and 91.07%, respectively. In the data set composed of air-door opening and closing and other mine production activity experimental data, the accuracy, precision, and recall rates of the air-door opening and closing identification are 89.61%, 90.31%, and 88.39%, respectively. [ABSTRACT FROM AUTHOR]

To achieve the full analytical potential of the streaming data from the internet of things, the interconnection of various data sources is needed. By definition, those sources are heterogeneous and their integration is not a trivial task. A common approach to exploit streaming sensor data potential is to use machine learning techniques for predictive analytics in a way that is agnostic to the domain knowledge. Such an approach can be easily integrated in various use cases. In this paper, we propose a novel framework for data fusion of a set of heterogeneous data streams. The proposed framework enriches streaming sensor data with the contextual and historical information relevant for describing the underlying processes. The final result of the framework is a feature vector, ready to be used in a machine learning algorithm. The framework has been applied to a cloud and to an edge device. In the latter case, incremental learning capabilities have been demonstrated. The reported results illustrate a significant improvement of data-driven models, applied to sensor streams. Beside higher accuracy of the models the platform offers easy setup and thus fast prototyping capabilities in real-world applications. [ABSTRACT FROM AUTHOR]

This paper introduces a novel semi-supervised tri-training classification algorithm based on regularized local discriminant embedding (RLDE) for hyperspectral imagery. In this algorithm, the RLDE method is used for optimal feature information extraction, to solve the problems of singular values and over-fitting, which are the main problems in the local discriminant embedding (LDE) and local Fisher discriminant analysis (LFDA) methods. An active learning method is then used to select the most useful and informative samples from the candidate set. In the experiments undertaken in this study, the three base classifiers were multinomial logistic regression (MLR), k-nearest neighbor (KNN), and random forest (RF). To confirm the effectiveness of the proposed RLDE method, experiments were conducted on two real hyperspectral datasets (Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) and Reflective Optics System Imaging Spectrometer (ROSIS)), and the proposed RLDE tri-training algorithm was compared with its counterparts of tri-training alone, LDE, and LFDA. The experiments confirmed that the proposed approach can effectively improve the classification accuracy for hyperspectral imagery. [ABSTRACT FROM AUTHOR]

Hyperspectral image classification is a hot issue in the field of remote sensing. It is possible to achieve high accuracy and strong generalization through a good classification method that is used to process image data. In this paper, an efficient hyperspectral image classification method based on improved Rotation Forest (ROF) is proposed. It is named ROF-KELM. Firstly, Non-negative matrix factorization(NMF) is used to do feature segmentation in order to get more effective data. Secondly, kernel extreme learning machine (KELM) is chosen as base classifier to improve the classification efficiency. The proposed method inherits the advantages of KELM and has an analytic solution to directly implement the multiclass classification. Then, Q-statistic is used to select base classifiers. Finally, the results are obtained by using the voting method. Three simulation examples, classification of AVIRIS image, ROSIS image and the UCI public data sets respectively, are conducted to demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Iris biometric detection provides contactless authentication, preventing the spread of COVID-19-like contagious diseases. However, these systems are prone to spoofing attacks attempted with the help of contact lenses, replayed video, and print attacks, making them vulnerable and unsafe. This paper proposes the iris liveness detection (ILD) method to mitigate spoofing attacks, taking global-level features of Thepade's sorted block truncation coding (TSBTC) and local-level features of the gray-level co-occurrence matrix (GLCM) of the iris image. Thepade's SBTC extracts global color texture content as features, and GLCM extracts local fine-texture details. The fusion of global and local content presentation may help distinguish between live and non-live iris samples. The fusion of Thepade's SBTC with GLCM features is considered in experimental validations of the proposed method. The features are used to train nine assorted machine learning classifiers, including naïve Bayes (NB), decision tree (J48), support vector machine (SVM), random forest (RF), multilayer perceptron (MLP), and ensembles (SVM + RF + NB, SVM + RF + RT, RF + SVM + MLP, J48 + RF + MLP) for ILD. Accuracy, precision, recall, and F-measure are used to evaluate the performance of the projected ILD variants. The experimentation was carried out on four standard benchmark datasets, and our proposed model showed improved results with the feature fusion approach. The proposed fusion approach gave 99.68% accuracy using the RF + J48 + MLP ensemble of classifiers, immediately followed by the RF algorithm, which gave 95.57%. The better capability of iris liveness detection will improve human–computer interaction and security in the cyber-physical space by improving person validation. [ABSTRACT FROM AUTHOR]

Critical technique errors are very often performed by patients in the use of Dry Powder Inhalers (DPIs) resulting in a reduction of the clinical efficiency of such medication. Those critical errors include: pure inhalation, non-arming of the device, no exhalation before or after inhalation, and non-holding of breath for 5–10 s between inhalation and exhalation. In this work, an audio-based classification method that assesses patient DPI user technique is presented by extracting the the non-silent audio segments and categorizing them into respiratory sounds. Twenty healthy and non-healthy volunteers used the same placebo inhaler (Bretaris Genuair Inhaler) in order to evaluate the performance of the algorithm. The audio-based method achieved an F1-score of 89.87% in classifying sound events (Actuation, Inhale, Button Press, and Exhale). The significance of the algorithm lies not just on automatic classification but on a post-processing step of peak detection that resulted in an improvement of 5.58% on the F1-score, reaching 94.85%. This method can provide a clinically accurate assessment of the patient's inhaler use without the supervision of a doctor. [ABSTRACT FROM AUTHOR]

In the malware detection process, obfuscated malicious codes cannot be efficiently and accurately detected solely in the dynamic or static feature space. Aiming at this problem, an integrative feature extraction algorithm based on simhash was proposed, which combines the static information e.g., API (Application Programming Interface) calls and dynamic information (such as file, registry and network behaviors) of malicious samples to form integrative features. The experiment extracts the integrative features of some static information and dynamic information, and then compares the classification, time and obfuscated-detection performance of the static, dynamic and integrated features, respectively, by using several common machine learning algorithms. The results show that the integrative features have better time performance than the static features, and better classification performance than the dynamic features, and almost the same obfuscated-detection performance as the dynamic features. This algorithm can provide some support for feature extraction of malware detection. [ABSTRACT FROM AUTHOR]

In view of the poor performance of traditional feature point detection methods in low-texture situations, we design a new self-supervised feature extraction network that can be applied to the visual odometer (VO) front-end feature extraction module based on the deep learning method. First, the network uses the feature pyramid structure to perform multi-scale feature fusion to obtain a feature map containing multi-scale information. Then, the feature map is passed through the position attention module and the channel attention module to obtain the feature dependency relationship of the spatial dimension and the channel dimension, respectively, and the weighted spatial feature map and the channel feature map are added element by element to enhance the feature representation. Finally, the weighted feature maps are trained for detectors and descriptors respectively. In addition, in order to improve the prediction accuracy of feature point locations and speed up the network convergence, we add a confidence loss term and a tolerance loss term to the loss functions of the detector and descriptor, respectively. The experiments show that our network achieves satisfactory performance under the Hpatches dataset and KITTI dataset, indicating the reliability of the network. [ABSTRACT FROM AUTHOR]

High accuracy and a real-time system are priorities in the development of a prosthetic hand. This study aimed to develop and evaluate a real-time embedded time-domain feature extraction and machine learning on a system on chip (SoC) Raspberry platform using a multi-thread algorithm to operate a prosthetic hand device. The contribution of this study is that the implementation of the multi-thread in the pattern recognition improves the accuracy and decreases the computation time in the SoC. In this study, ten healthy volunteers were involved. The EMG signal was collected by using two dry electrodes placed on the wrist flexor and wrist extensor muscles. To reduce the complexity, four time-domain features were applied to extract the EMG signal. Furthermore, these features were used as the input of the machine learning. The machine learning evaluated in this study were k-nearest neighbor (k-NN), Naive Bayes (NB), decision tree (DT), and support vector machine (SVM). In the SoC implementation, the data acquisition, feature extraction, machine learning, and motor control process were implemented using a multi-thread algorithm. After the evaluation, the result showed that the pairing of the MAV feature and machine learning DT resulted in higher accuracy among other combinations (98.41%) with a computation time of ~1 ms. The implementation of the multi-thread algorithm in the pattern recognition system resulted in significant impact on the time processing. [ABSTRACT FROM AUTHOR]

Developing cyber security is very necessary and has attracted considerable attention from academy and industry organizations worldwide. It is also very necessary to provide sustainable computing for the the Internet of Things (IoT). Machine learning techniques play a vital role in the cybersecurity of the IoT for intrusion detection and malicious identification. Thus, in this study, we develop new feature extraction and selection methods and for the IDS system using the advantages of the swarm intelligence (SI) algorithms. We design a feature extraction mechanism depending on the conventional neural networks (CNN). After that, we present an alternative feature selection (FS) approach using the recently developed SI algorithm, Aquila optimizer (AQU). Moreover, to assess the quality of the developed IDS approach, four well-known public datasets, CIC2017, NSL-KDD, BoT-IoT, and KDD99, were used. We also considered extensive comparisons to other optimization methods to verify the competitive performance of the developed method. The results show the high performance of the developed approach using different evaluation indicators. [ABSTRACT FROM AUTHOR]

A large number of intelligent models for masked face recognition (MFR) has been recently presented and applied in various fields, such as masked face tracking for people safety or secure authentication. Exceptional hazards such as pandemics and frauds have noticeably accelerated the abundance of relevant algorithm creation and sharing, which has introduced new challenges. Therefore, recognizing and authenticating people wearing masks will be a long-established research area, and more efficient methods are needed for real-time MFR. Machine learning has made progress in MFR and has significantly facilitated the intelligent process of detecting and authenticating persons with occluded faces. This survey organizes and reviews the recent works developed for MFR based on deep learning techniques, providing insights and thorough discussion on the development pipeline of MFR systems. State-of-the-art techniques are introduced according to the characteristics of deep network architectures and deep feature extraction strategies. The common benchmarking datasets and evaluation metrics used in the field of MFR are also discussed. Many challenges and promising research directions are highlighted. This comprehensive study considers a wide variety of recent approaches and achievements, aiming to shape a global view of the field of MFR. [ABSTRACT FROM AUTHOR]

The aim of this study was to find an efficient method to determine features that characterize octave illusion data. Specifically, this study compared the efficiency of several automatic feature selection methods for automatic feature extraction of the auditory steady-state responses (ASSR) data in brain activities to distinguish auditory octave illusion and nonillusion groups by the difference in ASSR amplitudes using machine learning. We compared univariate selection, recursive feature elimination, principal component analysis, and feature importance by testifying the results of feature selection methods by using several machine learning algorithms: linear regression, random forest, and support vector machine. The univariate selection with the SVM as the classification method showed the highest accuracy result, 75%, compared to 66.6% without using feature selection. The received results will be used for future work on the explanation of the mechanism behind the octave illusion phenomenon and creating an algorithm for automatic octave illusion classification. [ABSTRACT FROM AUTHOR]

The present work aims to demonstrate how machine learning (ML) techniques can be used for automatic feature detection and extraction in fluvial environments. The use of photogrammetry and machine learning algorithms has improved the understanding of both environmental and anthropic issues. The developed methodology was applied considering the acquisition of multiple photogrammetric images thanks to unmanned aerial vehicles (UAV) carrying multispectral cameras. These surveys were carried out in the Salbertrand area, along the Dora Riparia River, situated in Piedmont (Italy). The authors developed an algorithm able to identify and detect the water table contour concerning the landed areas: the automatic classification in ML found a valid identification of different patterns (water, gravel bars, vegetation, and ground classes) in specific hydraulic and geomatics conditions. Indeed, the RE+NIR data gave us a sharp rise in terms of accuracy by about 11% and 13.5% of F1-score average values in the testing point clouds compared to RGB data. The obtained results about the automatic classification led us to define a new procedure with precise validity conditions. [ABSTRACT FROM AUTHOR]

In human emotion estimation using an electroencephalogram (EEG) and heart rate variability (HRV), there are two main issues as far as we know. The first is that measurement devices for physiological signals are expensive and not easy to wear. The second is that unnecessary physiological indexes have not been removed, which is likely to decrease the accuracy of machine learning models. In this study, we used single-channel EEG sensor and photoplethysmography (PPG) sensor, which are inexpensive and easy to wear. We collected data from 25 participants (18 males and 7 females) and used a deep learning algorithm to construct an emotion classification model based on Arousal–Valence space using several feature combinations obtained from physiological indexes selected based on our criteria including our proposed feature selection methods. We then performed accuracy verification, applying a stratified 10-fold cross-validation method to the constructed models. The results showed that model accuracies are as high as 90% to 99% by applying the features selection methods we proposed, which suggests that a small number of physiological indexes, even from inexpensive sensors, can be used to construct an accurate emotion classification model if an appropriate feature selection method is applied. Our research results contribute to the improvement of an emotion classification model with a higher accuracy, less cost, and that is less time consuming, which has the potential to be further applied to various areas of applications. [ABSTRACT FROM AUTHOR]

Optical character recognition is gaining immense importance in the domain of deep learning. With each passing day, handwritten digits (0–9) data are increasing rapidly, and plenty of research has been conducted thus far. However, there is still a need to develop a robust model that can fetch useful information and investigate self-build handwritten digit data efficiently and effectively. The convolutional neural network (CNN) models incorporating a sigmoid activation function with a large number of derivatives have low efficiency in terms of feature extraction. Here, we designed a novel CNN model integrated with the extreme learning machine (ELM) algorithm. In this model, the sigmoid activation function is upgraded as the rectified linear unit (ReLU) activation function, and the CNN unit along with the ReLU activation function are used as a feature extractor. The ELM unit works as the image classifier, which makes the perfect symmetry for handwritten digit recognition. A deeplearning4j (DL4J) framework-based CNN-ELM model was developed and trained using the Modified National Institute of Standards and Technology (MNIST) database. Validation of the model was performed through self-build handwritten digits and USPS test datasets. Furthermore, we observed the variation of accuracies by adding various hidden layers in the architecture. Results reveal that the CNN-ELM-DL4J approach outperforms the conventional CNN models in terms of accuracy and computational time. [ABSTRACT FROM AUTHOR]

A nonlinear feature extraction-based approach using manifold learning algorithms is developed in order to improve the classification accuracy in an electronic tongue sensor array. The developed signal processing methodology is composed of four stages: data unfolding, scaling, feature extraction, and classification. This study aims to compare seven manifold learning algorithms: Isomap, Laplacian Eigenmaps, Locally Linear Embedding (LLE), modified LLE, Hessian LLE, Local Tangent Space Alignment (LTSA), and t-Distributed Stochastic Neighbor Embedding (t-SNE) to find the best classification accuracy in a multifrequency large-amplitude pulse voltammetry electronic tongue. A sensitivity study of the parameters of each manifold learning algorithm is also included. A data set of seven different aqueous matrices is used to validate the proposed data processing methodology. A leave-one-out cross validation was employed in 63 samples. The best accuracy ( 96.83 % ) was obtained when the methodology uses Mean-Centered Group Scaling (MCGS) for data normalization, the t-SNE algorithm for feature extraction, and k-nearest neighbors (kNN) as classifier. [ABSTRACT FROM AUTHOR]

Realization of navigation in virtual environments remains a challenge as it involves complex operating conditions. Decomposition of such complexity is attainable by fusion of sensors and machine learning techniques. Identifying the right combination of sensory information and the appropriate machine learning technique is a vital ingredient for translating physical actions to virtual movements. The contributions of our work include: (i) Synchronization of actions and movements using suitable multiple sensor units, and (ii) selection of the significant features and an appropriate algorithm to process them. This work proposes an innovative approach that allows users to move in virtual environments by simply moving their legs towards the desired direction. The necessary hardware includes only a smartphone that is strapped to the subjects' lower leg. Data from the gyroscope, accelerometer and campus sensors of the mobile device are transmitted to a PC where the movement is accurately identified using a combination of machine learning techniques. Once the desired movement is identified, the movement of the virtual avatar in the virtual environment is realized. After pre-processing the sensor data using the box plot outliers approach, it is observed that Artificial Neural Networks provided the highest movement identification accuracy of 84.2% on the training dataset and 84.1% on testing dataset. [ABSTRACT FROM AUTHOR]