Search

Walking robots are considered as a promising solution for locomotion across irregular or rough terrain. While wheeled or tracked robots require flat surface like roads or driveways, walking robots can adapt to almost any terrain type. However, overcoming diverse terrain obstacles still remains a challenging task even for multi-legged robots with a high number of degrees of freedom. Here, we present a novel method for obstacle overcoming for walking robots based on the use of tactile sensors and generative recurrent neural network for positional error prediction. By using tactile sensors positioned on the front side of the legs, we demonstrate that a robot is able to successfully overcome obstacles close to robots height in the terrains of different complexity. The proposed method can be used by any type of a legged machine and can be considered as a step toward more advanced walking robot locomotion in unstructured terrain and uncertain environment.

GF-3 satellite is China's first C-band multi-polarized synthetic aperture radar (SAR) satellite with the 1-meter resolution, which has been widely used in various fields. Road detection for GF-3 SAR images is an important part of the application of GF-3, especially in fields of map update, target recognition, and image matching. However, speckle appears in GF-3 SAR images due to coherent imaging system and it hinders the interpretation of images seriously. Especially the detection of weak roads under strong speckle background becomes extremely difficult. As a representative of multiscale geometric analysis (MGA) tool, shearlet has the optimal sparse representation feature and strong directional orientation, which can effectively capture edge and other anisotropic feature information, and can accurately describe the sparse characteristics of GF-3 SAR images. Based on shearlet, a method for detecting weak roads under strong speckle interference is proposed. Firstly, the Frost filter is used for despeckling. Secondly, shearlet is used for road detection. Finally, morphological operations are adopted to obtain the final result. Road detection experiments on various types of GF-3 SAR images demonstrate that, the proposed method can effectively overcome the interference of speckle, and completely and smoothly detect road information, which is very suitable for the detection of weak roads under strong speckle interference of GF-3 SAR images.

YouTube is a source of income for many people, and therefore a video’s popularity ultimately becomes the top priority for sustaining a steady income, provided that the popularity of videos remains the highest. Analysts and researchers use different algorithms and models to predict the maximum viewership of popular videos. This study predicts the popularity of such videos using the XGBoost model, considering features selection, fusion, min-max normalization and some precision parameters such as gamma, eta, learning_rate etc. The XGBoost gives 86% accuracy and 64% precision. Moreover, the Tuned XGboost also shows enhanced accuracy and precision. We have also analyzed the classification of unpopular videos for a comparison with our results. Finally, cross-validation methods are also used to evaluate certain combination of parameter’s values to validate our claims. Based on the obtained results, it can be said that our proposed models and techniques are very useful and can precisely and accurately predict the popularity of YouTube videos.

Chest diseases can be dangerous and deadly. They include many chest infections such as pneumonia, asthma, edema, and, lately, COVID-19. COVID-19 has many similar symptoms compared to pneumonia, such as breathing hardness and chest burden. However, it is a challenging task to differentiate COVID-19 from other chest diseases. Several related studies proposed a computer-aided COVID-19 detection system for the single-class COVID-19 detection, which may be misleading due to similar symptoms of other chest diseases. This paper proposes a framework for the detection of 15 types of chest diseases, including the COVID-19 disease, via a chest X-ray modality. Two-way classification is performed in proposed Framework. First, a deep learning-based convolutional neural network (CNN) architecture with a soft-max classifier is proposed. Second, transfer learning is applied using fully-connected layer of proposed CNN that extracted deep features. The deep features are fed to the classical Machine Learning (ML) classification methods. However, the proposed framework improves the accuracy for COVID-19 detection and increases the predictability rates for other chest diseases. The experimental results show that the proposed framework, when compared to other state-of-the-art models for diagnosing COVID-19 and other chest diseases, is more robust, and the results are promising.

A coal mine tunnel usually has a long-strip shape. Traditional single-sink cyber-physical-systems (CPS) cannot ensure that sensor nodes far from the entrance of a coal mine tunnel promptly and accurately transmit abnormal data from deep in the tunnel to the ground control center via the sink at the tunnel entrance. At the same time, it causes the “hot spot” problem easily around the Sink and limits the lifetime of the whole network. In order to solve these problems, this paper proposes a Multi-sink distributed power control algorithm (MSDPC-SRMS) suitable for a coal mine tunnel. This algorithm uses the Multiple Sink network structure and the idea of non-uniform cluster, combines the Multi-sink network with the clustering Voronoi scoping routing algorithm. It allocates the optimal transmission range and power for each sink, uses each sink as the cluster head and the network clustering is performed. The network topology is optimized on the basis of a good network coverage rate. Simulation results compared with CNP strategy show that the new algorithm exhibits superior connectivity, power consumption validity, clustering interference, and network performance, can effectively reduce the overall power consumption and prolong the lifetime of the network, thus ensure the monitoring data transmitted to the monitoring center rapidly and quickly.

Decision support systems become a very important part of our lives. Technologies make them applicable almost everywhere. These systems can simplify solutions to the numerous problems, speed up analysis of the results of medical research and contribute to the rapid classification of input patterns. Practical applications may not only contribute to efficient technology, but also improve important aspects of our lives. In this article, we propose a model of decision support system for speech processing. Proposed mechanism can be used in various applications, where voice sample can be evaluated by the use of the proposed methodology. The proposed solution is based on analysis of the speech signal through the developed intelligent technique in which the signal is processed by the composed mathematical transform cooperating with bio-inspired algorithm and spiking neural network to evaluate possible voice problems. A novelty of our idea is the approach to the topic from a different side, because graphical representations of audio signals and heuristic methods are composed for feature extraction. The results are discussed after extensive comparisons in terms of advantages and disadvantages of the proposed approach. As a part of the conducted research, we demonstrated which transformations and heuristic algorithms work better in the process of voice analysis.

Coreference resolution is a challenging part of natural language processing (NLP) with applications in machine translation, semantic search and other information retrieval, and decision support systems. Coreference resolution requires linguistic preprocessing and rich language resources for automatically identifying and resolving such expressions. Many rarer and under-resourced languages (such as Lithuanian) lack the required language resources and tools. We present a method for coreference resolution in Lithuanian language and its application for processing e-health records from a hospital reception. Our novelty is the ability to process coreferences with minimal linguistic resources, which is important in linguistic applications for rare and endangered languages. The experimental results show that coreference resolution is applicable to the development of NLP-powered online healthcare services in Lithuania.

We propose an algorithm to perform multi-focus image fusion and integrate a set of images acquired at different focus settings to a single uniformly focused image for visual inspection in assembly lines. The goal of image fusion is to integrate complementary image multi-view information from standard, low resolution assembly line camera into one new image, the quality of which could not be achieved using other methods such as direct digital photography. Our method is based on the image decomposition into Gaussian pyramids, generation of the Laplacian pyramids, and image reconstruction from the Laplacian pyramids. The main characteristics of the proposed method include good quality of integrated multi-focus image, and suitability for microscopy conveyor applications given movement of objects, different lighting conditions, and positional misalignments. We have evaluated our method using eight image quality metrics yielding good results (best results were obtained using the Tenengrad (TENG) method, reaching an accuracy of 0.982) with a low-cost camera and computationally efficient implementation.

Recognition of hand gestures (hand signals) is an active research area for human computer interaction with many possible applications. Automatic machine vision-based hand gesture interfaces for real-time applications require fast and extremely robust human, pose and hand detection, and gesture recognition. Attempting to recognize gestures performed by official referees in sports (such as basketball game) video places tough requirements on the image segmentation techniques. Here we propose an image segmentation technique based on the histogram of oriented gradients and local binary pattern (LBP) features, which allow recognizing the signals of basketball referee from recorded game videos and achieved an accuracy of 95.6% using LBP features and support vector machine for classification. Our results are relevant for real-time analysis of basketball game.

We present a method for the removal of movement artifacts from the recordings of electroencephalography (EEG) signals in the context of sports health. We use a smart wearable Internet of Things-based signal recording system to record physiological human signals [EEG, electrocardiography (ECG)] in real time. Then, the movement artifacts are removed using ECG as a reference signal and the baseline estimation and denoising with sparsity (BEADS) filter algorithm for trend removal. The parameters (cut-off frequency) of the BEADS filter are optimized with respect to the number of QRS complexes detected in the reference ECG signal. Next, surrogate movement signals are generated using a linear combination of intrinsic mode functions derived from the sample movement signals by the application of empirical mode decomposition. Surrogate signals are used to test the efficiency of the BEADS method for filtering the movement-contaminated EEG signals. We provide an analysis of the efficiency of the method, extracted movement artifacts and detrended EEG signals.

Methods for image registration, segmentation, and visualization of magnetic resonance imaging (MRI) data are used widely to help medical doctors in supporting diagnostics. The large amount and complexity of MRI data require looking for new methods that allow for efficient processing of this data. Here, we propose using the adaptive independent subspace analysis (AISA) method to discover meaningful electroencephalogram activity in the MRI scan data. The results of AISA (image subspaces) are analyzed using image texture analysis methods to calculate first order, gray-level co-occurrence matrix, gray-level size-zone matrix, gray-level run-length matrix, and neighboring gray-tone difference matrix features. The obtained feature space is mapped to the 2D space using the t-distributed stochastic neighbor embedding method. The classification results achieved using the k-nearest neighbor classifier with 10-fold cross-validation have achieved 94.7% of accuracy (and f-score of 0.9356) from the real autism spectrum disorder dataset.

Image matching is an important problem in computer vision and many technologies based on local descriptors have been developed. In this paper, we propose a novel local features descriptor based on an adaptive neighborhood and embedding Zernike moments. Instead of a fixed-size neighborhood, a size changeable neighborhood is introduced to detect the key-points and describe the features in the frame of Gaussian scale space. The radius is determined by the scale parameter of the key-point and the dominant direction is computed based on skew distribution fitting instead of the traditional eight-direction statistics. Then a 72-dimensional features vector based on a $3\times 3$ grid is presented. A 19-dimensional vector consists of Zernike moments is applied to achieve better rotation invariance and finally contributes to a 91-dimensional descriptor. The accuracy and efficiency of proposed descriptor for image matching are verified by several numerical experiments.

We propose an efficient mixture classification technique, which uses electroencephalography (EEG) signals for establishing a communication channel for the physically challenged or immobilized people, by the usage of the brain signals. In order to identify the emotion expressions by an immobilized person, we introduce a novel approach for emotion recognition based on the generalized mixture distribution model. The main benefit of utilizing this model is that it is an asymmetric distribution, which helps to extract the EEG signals, which are either in symmetric or asymmetric form. The skew Gaussian distribution helps to identify the small duration EEG signal sample and helps toward better recognition of emotions in both clean and noisy EEG signals. The proposed method is particularly well suited for the high variability of the EEG signal allowing the emotions to be identified appropriately. The features of the brain signals are extracted by using cepstral coefficients. The extracted features are classified into different emotions using mixture classification techniques. In order to validate the model, six mentally impaired subjects are considered in the age group of 60-68, and an 8-channel EEG signal is utilized to collect the EEG signals under audio-visual stimuli. The basic emotions considered in this study include happy, sad, neutral, and boredom and an average emotion recognition accuracy of 89% is achieved.

With the development of e-commerce, a large amount of personalized information is produced daily. To utilize diverse personalized information to improve recommendation accuracy, we propose a hybrid recommendation model based on users' ratings, reviews, and social data. Our model consists of six steps, review transformation, feature generation, community detection, model training, feature blending, and prediction and evaluation. Three groups of experiments are performed in this paper. Experiments A are used to identify the regression algorithm used in our model, Experiments B are used to identify the model to analyze review texts and the algorithm to detect social communities, and Experiments C compare our hybrid recommendation model with conventional recommendation models, such as probabilistic matrix factorization, UserKNN, ItemKNN, and social network-based models, such as socialMF and TrustSVD. The experiment results show that recommendation accuracy can be improved significantly with our hybrid model based on review texts and social communities.

Human activity recognition (HAR) plays a vital role in different real-world applications such as in tracking elderly activities for elderly care services, in assisted living environments, smart home interactions, healthcare monitoring applications, electronic games, and various human–computer interaction (HCI) applications, and is an essential part of the Internet of Healthcare Things (IoHT) services. However, the high dimensionality of the collected data from these applications has the largest influence on the quality of the HAR model. Therefore, in this paper, we propose an efficient HAR system using a lightweight feature selection (FS) method to enhance the HAR classification process. The developed FS method, called GBOGWO, aims to improve the performance of the Gradient-based optimizer (GBO) algorithm by using the operators of the grey wolf optimizer (GWO). First, GBOGWO is used to select the appropriate features, then, the support vector machine (SVM) is used to classify the activities. To assess the performance of GBOGWO, extensive experiments using well-known UCI-HAR and WISDM datasets were conducted. Overall outcomes show that GBOGWO improved the classification accuracy with an average accuracy of 98%.

Due to the rapid growth in artificial intelligence (AI) and deep learning (DL) approaches, the security and robustness of the deployed algorithms need to be guaranteed. The security susceptibility of the DL algorithms to adversarial examples has been widely acknowledged. The artificially created examples will lead to different instances negatively identified by the DL models that are humanly considered benign. Practical application in actual physical scenarios with adversarial threats shows their features. Thus, adversarial attacks and defense, including machine learning and its reliability, have drawn growing interest and, in recent years, has been a hot topic of research. We introduce a framework that provides a defensive model against the adversarial speckle-noise attack, the adversarial training, and a feature fusion strategy, which preserves the classification with correct labelling. We evaluate and analyze the adversarial attacks and defenses on the retinal fundus images for the Diabetic Retinopathy recognition problem, which is considered a state-of-the-art endeavor. Results obtained on the retinal fundus images, which are prone to adversarial attacks, are 99% accurate and prove that the proposed defensive model is robust.

The authenticity and integrity of medical images in telemedicine has to be protected. Robust reversible watermarking (RRW) algorithms provide copyright protection and the original images can be recovered at the receiver’s end. However, the existing algorithms have limitations in their ability to balance the tradeoff among robustness, imperceptibility, and embedded capacity. Some of them are even not completely reversible. Besides, most medical image watermarking algorithms are not designed for color images. To improve their performance in protecting medical color image information, we propose a novel RRW scheme based on the discrete wavelet transform (DWT). First, the DWT provides a robust solution. Second, the modification of the wavelet domain coefficient guarantees the changes of integer values in the spatial domain and ensures the reversibility of the watermarking scheme. Third, the embedding scheme makes full use of the characteristics of the original image and watermarking. This reduces the modification of the original image and ensures better imperceptibility. Lastly, the selection of the Zernike moments order for geometric correction is optimized to predict attack parameters more accurately by using less information. This enhances the robustness of the proposed scheme against geometric attacks such as rotation and scaling. The proposed scheme is robust against common and geometric attacks and has a high embedding capacity without obvious distortion of the image. The paper contributes towards improving the security of medical images in remote healthcare.

Manual diagnosis of skin cancer is time-consuming and expensive, therefore, it is essential to develop automated diagnostics methods with the ability to classify multiclass skin lesions with greater accuracy. We propose a fully automated approach for multiclass skin lesion segmentation and classification by using the most discriminant deep features. First, the input images are initially enhanced using local color-controlled histogram intensity values (LCcHIV). Next, saliency is estimated using a novel Deep Saliency Segmentation method, which uses a custom convolutional neural network (CNN) of ten layers. The generated heat map is converted into a binary image using a thresholding function. Next, the segmented color lesion images are used for feature extraction by a deep pre-trained CNN model. To avoid the curse of dimensionality, we implement an improved moth flame optimization (IMFO) algorithm to select the most discriminant features. The resultant features are fused using a multiset maximum correlation analysis (MMCA) and classified using the Kernel Extreme Learning Machine (KELM) classifier. The segmentation performance of the proposed methodology is analyzed on ISBI 2016, ISBI 2017, ISIC 2018, and PH2 datasets, achieving an accuracy of 95.38%, 95.79%, 92.69%, and 98.70%, respectively. The classification performance is evaluated on the HAM10000 dataset and achieved an accuracy of 90.67%. To prove the effectiveness of the proposed methods, we present a comparison with the state-of-the-art techniques.

Accurate price evaluation of real estate is beneficial for many parties involved in real estate business such as real estate companies, property owners, investors, banks, and financial institutes. Artificial Neural Networks (ANNs) have shown promising results in real estate price evaluation. However, the performance of ANNs greatly depends upon the settings of their hyperparameters. In this paper, we apply and optimize an ANN model for real estate price prediction in Helsinki, Finland. Optimization of the model is performed by fine-tuning hyper-parameters (such as activation functions, optimization algorithms, etc.) of the ANN architecture for higher accuracy using the Bayesian optimization algorithm. The results are evaluated using a variety of metrics (RMSE, MAE, R2) as well as illustrated graphically. The empirical analysis of the results shows that model optimization improved the performance on all metrics (reaching the relative mean error of 8.3%).

With increasing awareness of the advantages of game-based learning, there is a growing number of studies showing its application to both computer science education and sustainable development education. In this paper, we describe, with the example of the Eco JSity application, how both of these areas can be combined into a single tool. The presented evaluation results showed that despite incorporating the sustainable development topics into both the theme and mechanics of the educational game, both the usefulness and ease-of-use of the application were still assessed positively by students. We also identified three clusters of students depending on their attitude towards combining education on computer science and sustainable development. We concluded that game-based learning provided a capable means of incorporating sustainable development topics into teaching of unrelated technical skills.

Human posture detection allows the capture of the kinematic parameters of the human body, which is important for many applications, such as assisted living, healthcare, physical exercising and rehabilitation. This task can greatly benefit from recent development in deep learning and computer vision. In this paper, we propose a novel deep recurrent hierarchical network (DRHN) model based on MobileNetV2 that allows for greater flexibility by reducing or eliminating posture detection problems related to a limited visibility human torso in the frame, i.e., the occlusion problem. The DRHN network accepts the RGB-Depth frame sequences and produces a representation of semantically related posture states. We achieved 91.47% accuracy at 10 fps rate for sitting posture recognition.

Recently, there has been a huge rise in malware growth, which creates a significant security threat to organizations and individuals. Despite the incessant efforts of cybersecurity research to defend against malware threats, malware developers discover new ways to evade these defense techniques. Traditional static and dynamic analysis methods are ineffective in identifying new malware and pose high overhead in terms of memory and time. Typical machine learning approaches that train a classifier based on handcrafted features are also not sufficiently potent against these evasive techniques and require more efforts due to feature-engineering. Recent malware detectors indicate performance degradation due to class imbalance in malware datasets. To resolve these challenges, this work adopts a visualization-based method, where malware binaries are depicted as two-dimensional images and classified by a deep learning model. We propose an efficient malware detection system based on deep learning. The system uses a reweighted class-balanced loss function in the final classification layer of the DenseNet model to achieve significant performance improvements in classifying malware by handling imbalanced data issues. Comprehensive experiments performed on four benchmark malware datasets show that the proposed approach can detect new malware samples with higher accuracy (98.23% for the Malimg dataset, 98.46% for the BIG 2015 dataset, 98.21% for the MaleVis dataset, and 89.48% for the unseen Malicia dataset) and reduced false-positive rates when compared with conventional malware mitigation techniques while maintaining low computational time. The proposed malware detection solution is also reliable and effective against obfuscation attacks.

For people in developing countries, cassava is a major source of calories and carbohydrates. However, Cassava Mosaic Disease (CMD) has become a major cause of concern among farmers in sub-Saharan Africa countries, which rely on cassava for both business and local consumption. The article proposes a novel deep residual convolution neural network (DRNN) for CMD detection in cassava leaf images. With the aid of distinct block processing, we can counterbalance the imbalanced image dataset of the cassava diseases and increase the number of images available for training and testing. Moreover, we adjust low contrast using Gamma correction and decorrelation stretching to enhance the color separation of an image with significant band-to-band correlation. Experimental results demonstrate that using a balanced dataset of images increases the accuracy of classification. The proposed DRNN model outperforms the plain convolutional neural network (PCNN) by a significant margin of 9.25% on the Cassava Disease Dataset from Kaggle.

In this Internet age, there are increasingly many threats to the security and safety of users daily. One of such threats is malicious software otherwise known as malware (ransomware, Trojans, viruses, etc.). The effect of this threat can lead to loss or malicious replacement of important information (such as bank account details, etc.). Malware creators have been able to bypass traditional methods of malware detection, which can be time-consuming and unreliable for unknown malware. This motivates the need for intelligent ways to detect malware, especially new malware which have not been evaluated or studied before. Machine learning provides an intelligent way to detect malware and comprises two stages: feature extraction and classification. This study suggests an ensemble learning-based method for malware detection. The base stage classification is done by a stacked ensemble of fully-connected and one-dimensional convolutional neural networks (CNNs), whereas the end-stage classification is done by a machine learning algorithm. For a meta-learner, we analyzed and compared 15 machine learning classifiers. For comparison, five machine learning algorithms were used: naïve Bayes, decision tree, random forest, gradient boosting, and AdaBoosting. The results of experiments made on the Windows Portable Executable (PE) malware dataset are presented. The best results were obtained by an ensemble of seven neural networks and the ExtraTrees classifier as a final-stage classifier.

Globally, breast cancer is one of the most significant causes of death among women. Early detection accompanied by prompt treatment can reduce the risk of death due to breast cancer. Currently, machine learning in cloud computing plays a pivotal role in disease diagnosis, but predominantly among the people living in remote areas where medical facilities are scarce. Diagnosis systems based on machine learning act as secondary readers and assist radiologists in the proper diagnosis of diseases, whereas cloud-based systems can support telehealth services and remote diagnostics. Techniques based on artificial neural networks (ANN) have attracted many researchers to explore their capability for disease diagnosis. Extreme learning machine (ELM) is one of the variants of ANN that has a huge potential for solving various classification problems. The framework proposed in this paper amalgamates three research domains: Firstly, ELM is applied for the diagnosis of breast cancer. Secondly, to eliminate insignificant features, the gain ratio feature selection method is employed. Lastly, a cloud computing-based system for remote diagnosis of breast cancer using ELM is proposed. The performance of the cloud-based ELM is compared with some state-of-the-art technologies for disease diagnosis. The results achieved on the Wisconsin Diagnostic Breast Cancer (WBCD) dataset indicate that the cloud-based ELM technique outperforms other results. The best performance results of ELM were found for both the standalone and cloud environments, which were compared. The important findings of the experimental results indicate that the accuracy achieved is 0.9868, the recall is 0.9130, the precision is 0.9054, and the F1-score is 0.8129.

Since the emergence of COVID-19, thousands of people undergo chest X-ray and computed tomography scan for its screening on everyday basis. This has increased the workload on radiologists, and a number of cases are in backlog. This is not only the case for COVID-19, but for the other abnormalities needing radiological diagnosis as well. In this work, we present an automated technique for rapid diagnosis of COVID-19 on computed tomography images. The proposed technique consists of four primary steps: (1) data collection and normalization, (2) extraction of the relevant features, (3) selection of the most optimal features and (4) feature classification. In the data collection step, we collect data for several patients from a public domain website, and perform preprocessing, which includes image resizing. In the successive step, we apply discrete wavelet transform and extended segmentation-based fractal texture analysis methods for extracting the relevant features. This is followed by application of an entropy controlled genetic algorithm for selection of the best features from each feature type, which are combined using a serial approach. In the final phase, the best features are subjected to various classifiers for the diagnosis. The proposed framework, when augmented with the Naive Bayes classifier, yields the best accuracy of 92.6%. The simulation results are supported by a detailed statistical analysis as a proof of concept.

There still exists a knowledge gap in the field of computer vision in respect of posture prediction and deviation evaluation is an important metric for various medical applications, that require posture abnormality quantization. Our paper proposes a deep heuristic neural network architecture, using BlazePose as a backbone, that is capable of reconstructing users skeleton from a real-time monocular video feed, using which we are able to evaluate the subjects performed exercise and measure the deviation from expected values. The proposed heuristics are able to identify and evaluate most of the abnormalities, with the highest indicator of postural issues being the spinal deviation accounting for 95%. Additional evaluation of real-time performance has shown that our method is capable of maintaining 23-ms response times, making it applicable to real-time applications.

Multimodal image fusion combines the complementary information of multimodality images into a single image that preserves the information of all the source images. This paper proposes a multimodal image fusion method situated on image enhancement, edge detection, multiscale sliding window, and image matting to obtain the detailed region information of the input images. In the proposed system, firstly the multimodality input images are rectified via a contrast enhancement method through which the intensity distribution is refined for clear vision. The spatial gradient edge detection method is utilized for separating the edge information from the contrast-enhanced images. These edges are then used by a multiscale sliding window method to provide global and local activity level maps. These activity maps further generate trimap and decision maps. Finally, by employing the improved decision maps and fusion rule the fused image is acquired.

All civil infrastructure units demand regular inspection to avoid functional and structural damages. Periodic examinations are in accordance with classification society’s standards which contain both non-destructive tests and visual surveys, to search structural damage, reliability, cracks, thickness measurement, and Water dripping generally documented by manually or with measurements tape. But, it is very hard to search cracks by visually monitoring much larger structures. Hence, the advent of crack detecting and monitoring systems has been a major issue. In this proposed study a crack detection algorithm in reference to digital image processing technology is suggested. Obtaining information of a surface crack by using an image pre-processing pipeline and to estimate the failure is proposed which helps and detects the structure cracks and body health information. It provides the identification system more portable and integrated, estimates the crack more precisely and reduction in expenditure as well. The proposed algorithm accuracy is 93.8% as compared to the traditional and recent work.

Facial symmetry is a key component in quantifying the perception of beauty. In this paper, we propose a set of facial features computed from facial landmarks which can be extracted at a low computational cost. We quantitatively evaluated the proposed features for predicting perceived attractiveness from human portraits on four benchmark datasets (SCUT-FBP, SCUT-FBP5500, FACES and Chicago Face Database). Experimental results showed that the performance of the proposed features is comparable to those extracted from a set with much denser facial landmarks. The computation of facial features was also implemented as an augmented reality (AR) app developed on Android OS. The app overlays four types of measurements and guidelines over a live video stream, while the facial measurements are computed from the tracked facial landmarks at run time. The developed app can be used to assist plastic surgeons in assessing facial symmetry when planning reconstructive facial surgeries.

We propose a deep learning method based on the Region Based Convolutional Neural Networks (R-CNN) architecture for the evaluation of sperm head motility in human semen videos. The neural network performs the segmentation of sperm heads, while the proposed central coordinate tracking algorithm allows us to calculate the movement speed of sperm heads. We have achieved 91.77% (95% CI, 91.11%&ndash, 92.43%) accuracy of sperm head detection on the VISEM (A Multimodal Video Dataset of Human Spermatozoa) sperm sample video dataset. The mean absolute error (MAE) of sperm head vitality prediction was 2.92 (95% CI, 2.46&ndash, 3.37), while the Pearson correlation between actual and predicted sperm head vitality was 0.969. The results of the experiments presented below will show the applicability of the proposed method to be used in automated artificial insemination workflow.

Documents are stored in a digital form across several organizations. Printing this amount of data and placing it into folders instead of storing digitally is against the practical, economical, and ecological perspective. An efficient way of retrieving data from digitally stored documents is also required. This article presents a real-time supervised learning technique for document classification based on deep convolutional neural network (DCNN), which aims to reduce the impact of adverse document image issues such as signatures, marks, logo, and handwritten notes. The proposed technique&rsquo, s major steps include data augmentation, feature extraction using pre-trained neural network models, feature fusion, and feature selection. We propose a novel data augmentation technique, which normalizes the imbalanced dataset using the secondary dataset RVL-CDIP. The DCNN features are extracted using the VGG19 and AlexNet networks. The extracted features are fused, and the fused feature vector is optimized by applying a Pearson correlation coefficient-based technique to select the optimized features while removing the redundant features. The proposed technique is tested on the Tobacco3482 dataset, which gives a classification accuracy of 93.1% using a cubic support vector machine classifier, proving the validity of the proposed technique.

The streamflow of the upper reaches of the Yangtze River exhibits different timing and periodicity characteristics in different quarters and months of the year, which makes it difficult to predict. Existing sliding window-based methods usually use a fixed-size window, for which the window size selection is random, resulting in large errors. This paper proposes a dynamic sliding window method that reflects the different timing and periodicity characteristics of the streamflow in different months of the year. Multiple datasets of different months are generated using a dynamic window at first, then the long-short term memory (LSTM) is used to select the optimal window, and finally, the dataset of the optimal window size is used for verification. The proposed method was tested using the hydrological data of Zhutuo Hydrological Station (China). A comparison between the flow prediction data and the measured data shows that the prediction method based on a dynamic sliding window LSTM is more accurate by 8.63% and 3.85% than the prediction method based on fixed window LSTM and the dynamic sliding window back-propagation neural network, respectively. This method can be generally used for the time series data prediction with different periodic characteristics.

This paper presents a novel design of a Multiple Input Multiple Output (MIMO) antenna system for next generation sub 6 GHz 5G and beyond mobile terminals. The proposed system is composed of a main board and two side boards. To make the design cost-effective, FR4 is used as a substrate. The design is based on a unit monopole antenna etched at the side substrate. The single element is resonating at 3.5 GHz attaining a 10 dB bandwidth of 200 MHz and a 6 dB bandwidth of 400 MHz. The single element is then transformed into an MIMO array of 8-elements with an overall dimension of 150 mm &times, 75 mm &times, 7 mm, providing pattern diversity characteristics and isolation better than &minus, 12 dB for any two radiating elements. A number of studies such as effects of human hand on the system that includes single hand mode and dual mode scenarios and the effects of Liquid Crystal Display (LCD) over the principal performance parameters of the system are presented. The envelop correlation coefficient (ECC) is computed for all the scenarios and it is found that ECC is less than 0.1 for any case and maximum channel capacity is 38.5 bps/Hz within the band of interest. The main advantage of the proposed design over available designs in the literature is that almost all of the main substrate is empty providing wide space for different sensors, systems, and mobile technology components. A brief literature comparison of the proposed system is also presented. To validate the proposed model, a prototype is fabricated and results are presented. This design can be applied on higher frequencies to future micromachines for on chip communications using same theocratical approach as the space for higher frequencies in mmwave spectrum has been reserved. The simulated results are in an excellent agreement with the measured results. All the main performance parameters of the design are calculated and compared with the measured results wherever possible.

Scheduling can be described as a decision-making process. It is applied in various applications, such as manufacturing, airports, and information processing systems. More so, the presence of symmetry is common in certain types of scheduling problems. There are three types of parallel machine scheduling problems (PMSP): uniform, identical, and unrelated parallel machine scheduling problems (UPMSPs). Recently, UPMSPs with setup time had attracted more attention due to its applications in different industries and services. In this study, we present an efficient method to address the UPMSPs while using a modified harris hawks optimizer (HHO). The new method, called MHHO, uses the salp swarm algorithm (SSA) as a local search for HHO in order to enhance its performance and to decrease its computation time. To test the performance of MHHO, several experiments are implemented using small and large problem instances. Moreover, the proposed method is compared to several state-of-art approaches used for UPMSPs. The MHHO shows better performance in both small and large problem cases.

The segmentation of high-resolution (HR) remote sensing images is very important in modern society, especially in the fields of industry, agriculture and urban modelling. Through the neural network, the machine can effectively and accurately extract the surface feature information. However, using the traditional deep learning methods requires plentiful efforts in order to find a robust architecture. In this paper, we introduce a neural network architecture search (NAS) method, called NAS-HRIS, which can automatically search neural network architecture on the dataset. The proposed method embeds a directed acyclic graph (DAG) into the search space and designs the differentiable searching process, which enables it to learn an end-to-end searching rule by using gradient descent optimization. It uses the Gumbel-Max trick to provide an efficient way when drawing samples from a non-continuous probability distribution, and it improves the efficiency of searching and reduces the memory consumption. Compared with other NAS, NAS-HRIS consumes less GPU memory without reducing the accuracy, which corresponds to a large amount of HR remote sensing imagery data. We have carried out experiments on the WHUBuilding dataset and achieved 90.44% MIoU. In order to fully demonstrate the feasibility of the method, we made a new urban Beijing Building dataset, and conducted experiments on satellite images and non-single source images, achieving better results than SegNet, U-Net and Deeplab v3+ models, while the computational complexity of our network architecture is much smaller.

Manual identification of brain tumors is an error-prone and tedious process for radiologists, therefore, it is crucial to adopt an automated system. The binary classification process, such as malignant or benign is relatively trivial, whereas, the multimodal brain tumors classification (T1, T2, T1CE, and Flair) is a challenging task for radiologists. Here, we present an automated multimodal classification method using deep learning for brain tumor type classification. The proposed method consists of five core steps. In the first step, the linear contrast stretching is employed using edge-based histogram equalization and discrete cosine transform (DCT). In the second step, deep learning feature extraction is performed. By utilizing transfer learning, two pre-trained convolutional neural network (CNN) models, namely VGG16 and VGG19, were used for feature extraction. In the third step, a correntropy-based joint learning approach was implemented along with the extreme learning machine (ELM) for the selection of best features. In the fourth step, the partial least square (PLS)-based robust covariant features were fused in one matrix. The combined matrix was fed to ELM for final classification. The proposed method was validated on the BraTS datasets and an accuracy of 97.8%, 96.9%, 92.5% for BraTs2015, BraTs2017, and BraTs2018, respectively, was achieved.

Video stitching technology provides an effective solution for a wide viewing angle monitoring mode for industrial applications. At present, the observation angle of a single camera is limited, and the monitoring network composed of multiple cameras will have many overlapping images captured. Monitoring surveillance cameras can cause the problems of viewing fatigue and low video utilization rate of involved personnel. In addition, current video stitching technology has poor adaptability and real-time performance. We propose an effective hybrid image feature detection method for fast video stitching of mine surveillance video using the effective information of the surveillance video captured from multiple cameras in the actual conditions in the industrial coal mine. The method integrates the Moravec corner point detection and the scale-invariant feature transform (SIFT) feature extractor. After feature extraction, the nearest neighbor method and the random sampling consistency (RANSAC) algorithm are used to register the video frames. The proposed method reduces the image stitching time and solves the problem of feature re-extraction due to the change of observation angle, thus optimizing the entire video stitching process. The experimental results on the real-world underground mine videos show that the optimized stitching method can stitch videos at a speed of 21 fps, effectively meeting the real-time requirement, while the stitching effect has a good stability and applicability in real-world conditions.

As the number of internet users increases so does the number of malicious attacks using malware. The detection of malicious code is becoming critical, and the existing approaches need to be improved. Here, we propose a feature fusion method to combine the features extracted from pre-trained AlexNet and Inception-v3 deep neural networks with features attained using segmentation-based fractal texture analysis (SFTA) of images representing the malware code. In this work, we use distinctive pre-trained models (AlexNet and Inception-V3) for feature extraction. The purpose of deep convolutional neural network (CNN) feature extraction from two models is to improve the malware classifier accuracy, because both models have characteristics and qualities to extract different features. This technique produces a fusion of features to build a multimodal representation of malicious code that can be used to classify the grayscale images, separating the malware into 25 malware classes. The features that are extracted from malware images are then classified using different variants of support vector machine (SVM), k-nearest neighbor (KNN), decision tree (DT), and other classifiers. To improve the classification results, we also adopted data augmentation based on affine image transforms. The presented method is evaluated on a Malimg malware image dataset, achieving an accuracy of 99.3%, which makes it the best among the competing approaches.

Tuberculosis (TB) is is an infectious disease that generally attacks the lungs and causes death for millions of people annually. Chest radiography and deep-learning-based image segmentation techniques can be utilized for TB diagnostics. Convolutional Neural Networks (CNNs) has shown advantages in medical image recognition applications as powerful models to extract informative features from images. Here, we present a novel hybrid method for efficient classification of chest X-ray images. First, the features are extracted from chest X-ray images using MobileNet, a CNN model, which was previously trained on the ImageNet dataset. Then, to determine which of these features are the most relevant, we apply the Artificial Ecosystem-based Optimization (AEO) algorithm as a feature selector. The proposed method is applied to two public benchmark datasets (Shenzhen and Dataset 2) and allows them to achieve high performance and reduced computational time. It selected successfully only the best 25 and 19 (for Shenzhen and Dataset 2, respectively) features out of about 50,000 features extracted with MobileNet, while improving the classification accuracy (90.2% for Shenzen dataset and 94.1% for Dataset 2). The proposed approach outperforms other deep learning methods, while the results are the best compared to other recently published works on both datasets.

Since the epidemic outbreak in early months of 2020 the spread of COVID-19 has grown rapidly in most countries and regions across the World. Because of that, SARS-CoV-2 was declared as a Public Health Emergency of International Concern (PHEIC) on January 30, 2020, by The World Health Organization (WHO). That’s why many scientists are working on new methods to reduce further growth of new cases and, by intelligent patients allocation, reduce number of patients per doctor, what can lead to more successful treatments. However to properly manage the COVID-19 spread there is a need for real-time prediction models which can reliably support various decisions both at national and international level. The problem in developing such system is the lack of general knowledge how the virus spreads and what would be the number of cases each day. Therefore prediction model must be able to conclude the situation from past data in the way that results will show a future trend and will possibly closely relate to the real numbers. In our opinion Artificial Intelligence gives a possibility to do it. In this article we present a model which can work as a part of an online system as a real-time predictor to help in estimation of COVID-19 spread. This prediction model is developed using Artificial Neural Networks (ANN) to estimate the future situation by the use of geo-location and numerical data from past 2 weeks. The results of our model are confirmed by comparing them with real data and, during our research the model was correctly predicting the trend and very closely matching the numbers of new cases in each day.

We adopt Bidirectional Long Short-Term Memory (BiLSTM) neural network and Wavelet Scattering Transform with Support Vector Machine (WST-SVM) classifier for detecting speech impairments of patients at the early stage of central nervous system disorders (CNSD). The study includes 339 voice samples collected from 15 subjects: 7 patients with early stage CNSD (3 Huntington, 1 Parkinson, 1 cerebral palsy, 1 post stroke, 1 early dementia), other 8 subjects were healthy. Speech data is collected using voice recorder from Neural Impairment Test Suite (NITS) mobile app. Features are extracted from pitch contours, Mel-frequency cepstral coefficients (MFCC), Gammatone cepstral coefficients (GTCC), Gabor (analytic Morlet) wavelet and auditory spectrograms. 94.50% (BiLSTM) and 96.3% (WST-SVM) accuracy is achieved for solving healthy vs. impaired classification problem. The developed method can be applied for automated CNSD patient health state monitoring and clinical decision support systems as well as a part of Internet of Medical Things (IoMT).

Infrared image recognition technology can work day and night and has a long detection distance. However, the infrared objects have less prior information and external factors in the real-world environment easily interfere with them. Therefore, infrared object classification is a very challenging research area. Manifold learning can be used to improve the classification accuracy of infrared images in the manifold space. In this article, we propose a novel manifold learning algorithm for infrared object detection and classification. First, a manifold space is constructed with each pixel of the infrared object image as a dimension. Infrared images are represented as data points in this constructed manifold space. Next, we simulate the probability distribution information of infrared data points with the Gaussian distribution in the manifold space. Then, based on the Gaussian distribution information in the manifold space, the distribution characteristics of the data points of the infrared image in the low-dimensional space are derived. The proposed algorithm uses the Kullback-Leibler (KL) divergence to minimize the loss function between two symmetrical distributions, and finally completes the classification in the low-dimensional manifold space. The efficiency of the algorithm is validated on two public infrared image data sets. The experiments show that the proposed method has a 97.46% classification accuracy and competitive speed in regards to the analyzed data sets.

State-of-the-art intelligent versatile applications provoke the usage of full 3D, depth-based streams, especially in the scenarios of intelligent remote control and communications, where virtual and augmented reality will soon become outdated and are forecasted to be replaced by point cloud streams providing explorable 3D environments of communication and industrial data. One of the most novel approaches employed in modern object reconstruction methods is to use a priori knowledge of the objects that are being reconstructed. Our approach is different as we strive to reconstruct a 3D object within much more difficult scenarios of limited data availability. Data stream is often limited by insufficient depth camera coverage and, as a result, the objects are occluded and data is lost. Our proposed hybrid artificial neural network modifications have improved the reconstruction results by 8.53% which allows us for much more precise filling of occluded object sides and reduction of noise during the process. Furthermore, the addition of object segmentation masks and the individual object instance classification is a leap forward towards a general-purpose scene reconstruction as opposed to a single object reconstruction task due to the ability to mask out overlapping object instances and using only masked object area in the reconstruction process.

Meander structures are highly relevant in the Internet-of-Things (IoT) communication systems, their miniaturization remains as one of the key design issues. Meander structures allow to decrease the size of the IoT device, while maintaining the same operating parameters of the IoT device. Meander structures can also work as the delay systems, which can be used for the delay and synchronization of signals in IoT devices. The design procedure of the meander delay systems is time-consuming and cumbersome because of the complexity of the numerical and analytical methods employed during the design process. New methods, which will accelerate the synthesis procedure of the meander delay systems, should be investigated. This is especially relevant when the procedure of synthesis must be repeated many times until the appropriate configuration of the IoT device is found. We present the procedure of synthesis of the meander delay system using the Pareto-optimal multilayer perceptron network and multiple linear regression model with the M5 descriptor. The prediction results are compared with results, which were obtained using the commercial Sonnet© software package and with the results of physical experiment. The difference between the experimentally achieved and predicted results did not exceed 1.53 %. Moreover, the prediction of parameters of the meander delay system allowed to speed up the procedure of synthesis multiple times from hours to only 2.3 s.

Image recognition by deep learning usually requires many sample images to train. In case of a small number of images available for training, data augmentation techniques should be applied. Here we propose a novel image augmentation technique based on a random permutation of coefficients of within-class principal components obtained after applying Principal Component Analysis (PCA). After reconstruction, newly generated surrogate images are employed to train a deep network. In this study, we demonstrated the applicability of our approach on training a custom convolutional neural network using the CIFAR-10 image dataset. The experimental results show an improvement in terms of classification accuracy and classification ambiguity.

Neural networks (NNs) have recently achieved significant performance gains for time series forecasting. However, they require large amounts of data to train. Data augmentation techniques have been suggested to improve the network training performance. Here, we adopt Empirical Mode Decomposition (EMD) as a data augmentation technique. The intrinsic mode functions (IMFs) produced by EMD are recombined with different weights to obtain surrogate data series, which are used to train a neural network for forecasting. We use M4 time series dataset and a custom nonlinear auto-regressive network with exogenous inputs (NARX) for the validation of the proposed method. The experimental results show an improvement in forecasting accuracy over a baseline method when using EMD based data augmentation.

Improving safety management and risk evaluation methods is important for the global mining industry, which is the backbone of the industrial development of our society. To prevent any accidental loss or harm to human life and property, a safety risk assessment method is needed to perform the continuous risk assessment of mines. Based on the requirements of mine safety evaluation, this paper proposes the Pb-Zn mine safety risk evaluation model based on the fuzzy-grey correlation analysis method. The model is compared with the risk assessment model based on the fuzzy TOPSIS method. Through the experiments, our results demonstrate that the proposed fuzzy-grey correlation model is more sensitive to risk and has less effect on the evaluation results under different scoring attitudes (cautious, rational, and relaxed).

Smartphones are indispensable helpers of people in their daily routine, including functions that serve the disabled or aged people finding their outdoor location. This paper presents an experimental investigation of the indoor positioning algorithms based on the signal strength received from the Bluetooth Low Energy (BLE) beacons. We have implemented and compared several positioning algorithms (Proximity Localization, Centroid Localization, Weighted Centroid Localization, Weight-Compensated Weighted Centroid Localization Based on RSSI, Fingerprinting and Trilateration Localization). We also proposed and implemented a fuzzy logic based scheme to select the most fitting algorithm depending upon the size of the room, the number of beacons available and the strength of the signal. We have evaluated our scheme in real-world conditions (office building). The experimental results show that the algorithm of fingerprinting localization is the most suitable one. Finally, we propose a fuzzy logic system for the selection of an indoor localization algorithm based on the size of the room, the number of available beacons, and the strength of RSSI signal.