Search

Generating natural language description for visual content is a technique for describing the content available in the image(s). It requires knowledge of both the domains of computer vision and natural language processing. For this, various models with different approaches are suggested. One of them is encoder-decoder-based description generation. Existing papers used only objects for descriptions, but the relationship between them is equally essential, requiring context information. Which required techniques like Long Short-Term Memory (LSTM). This paper proposes an encoder-decoder-based methodology to generate human-like textual descriptions. Dense-LSTM is presented for better description as a decoder with a modified VGG19 encoder to capture information to describe the scene. Standard datasets Flickr8K and Flickr30k are used for testing and training purposes. BLEU (Bilingual Evaluation Understudy) score is used to evaluate the generated text. For the proposed model, a GUI (Graphical User Interface) is developed, which produces the audio description of the output received and provides an interface for searching the related visual content and query-based search. [ABSTRACT FROM AUTHOR]

Because of the two sequenced methods stated above, SG and AMP, are being used in different ways, present a deep learning methodology for taxonomic categorization of the metagenomic information which could be utilized for either. To place the suggested pipeline to a trial, 1000 16 S full-length genomes were used to generate either SG or AMP short-reads. Then, to map sequencing as matrices into such a number space, used a k-mer model. Our analysis of the existing approaches revealed several drawbacks, including limited ability to handle complex hierarchical representations of data and suboptimal feature extraction from grid-like structures. To overcome these limitations, we introduce DBNs for feature learning and dimensionality reduction, and CNNs for efficient processing of grid-like metagenomic data. Finally, a training set for every taxon was obtained by training two distinct deep learning constructions, specifically deep belief network (DBN) and convolutional neural network (CNN). This examined the proposed methodology to determine the best factor that determines and compared findings to the classification abilities offered by the RDP classifier, a standard classifier for bacterium identification. These designs outperform using RDP classifiers at every taxonomic level. So, at the genetic level, for example, both CNN and DBN achieved 91.4% accuracy using AMP short-reads, but the RDP classifier achieved 83.9% with the same information. This paper, suggested a classification method for 16 S short-read sequences created on k-mer representations and a deep learning structure, that every taxon creates a classification method. The experimental findings validate the suggested pipelines as a realistic strategy for classifying bacterium samples; as a result, the technique might be included in the most commonly used tools for the metagenomic research. According to the outcomes, it could be utilized to effectively classify either SG or AMP information. [ABSTRACT FROM AUTHOR]

Differential protection of power transformers, as the fundamental protection, plays an important role in power system reliability and security. The main challenge in differential protection is discrimination between internal faults of power transformers and inrush current. Development of differential protection, especially discrimination between internal faults from other disturbances, have been a favorite subject in power system protection field over decades. Traditional methods proposed so far have several shortcomings: i) high computational burden, ii) sensitivity to noise, iii) being influenced by predefined threshold value/additional parameters/different models at varying ambient conditions, and iv) dependence on handcrafted or spectral analysis to extract features. Deep neural networks (DNN) is selected as the potential solution in this paper, which is able to capture the hierarchical features of a half-cycle of raw data. This paper proposes convolutional neural networks (CNN), in which batch normalization and scaled exponential linear unit (SELU) are merged to enhance differential protection performance. In order to generalize the CNN-based differential protection, several external factors, i.e. the compensation error of current transformer (CT) saturation, series compensated line, and superconducting fault current limiter (SFCL) are conducted to verify the reliability of the proposed method through different reliability metrics. The simulation and experimental results are assessed to show high reliability and the speed of the proposed method. [ABSTRACT FROM AUTHOR]

Accurate SARS-CoV-2 screening is made possible by automated Computer-Aided Diagnosis (CAD) which reduces the stress on healthcare systems. Since Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) is highly contagious, the transition chain can be broken through an early diagnosis by clinical knowledge and Artificial Intelligence (AI). Manual findings are time and labor-intensive. Even if Reverse Transcription-Polymerase Chain Reaction (RT-PCR) delivers quick findings, Chest X-ray (CXR) imaging is still a more trustworthy tool for disease classification and assessment. Several studies have been conducted using Deep Learning (DL) algorithms for COVID-19 detection. One of the biggest challenges in modernizing healthcare is extracting useful data from high-dimensional, heterogeneous, and complex biological data. Intending to introduce an automated COVID-19 diagnosis model, this paper develops a proficient optimization model that enhances the classification performance with better accuracy. The input images are initially pre-processed with an image filtering approach for noise removal and data augmentation to extend the dataset. Secondly, the images are segmented via U-Net and are given to classification using the Fused U-Net Convolutional Neural Network (FUCNN) model. Here, the performance of U-Net is enhanced through the modified Moth Flame Optimization (MFO) algorithm named Chaotic System-based MFO (CSMFO) by optimizing the weights of U-Net. The significance of the implemented model is confirmed over a comparative evaluation with the state-of-the-art models. Specifically, the proposed CSMFO-FUCNN attained 98.45% of accuracy, 98.63% of sensitivity, 98.98% of specificity, and 98.98% of precision. [ABSTRACT FROM AUTHOR]

Nowadays, automatic human activity recognition from video images is necessary for monitoring applications and caring for disabled people. The use of surveillance cameras and the processing of the obtained images leads to the achievement of a smart, accurate system for the recognition of human behavior. Since human detection in different scenes is associated with many challenges, several approaches have been implemented to detect human activity from video image processing. Due to the complexity of human activities, background noises and other factors affect the detection. For the solution of these problems, two deep learning-based algorithms have been described in the current article. According to the convolutional neural networks, the LSTM + CNN method and the 3D CNN method have been used to recognize the human activities in the images of the video. Each algorithm is explained and analyzed in detail. The experiments designed in this paper are performed by two datasets: the HMDB-51 dataset and the UCF101 dataset. In the HMDB-51 dataset, the highest obtained accuracy for CNN + LSTM method was equal to 70.2 and for method 3D CNN equal to 54.4. In the UCF101 dataset, the highest obtained accuracy for CNN + LSTM method was equal to 95.1 and for method 3D CNN equal to 90.8. [ABSTRACT FROM AUTHOR]

Pavement crack assessment is an important indicator for evaluating road health. However, due to the dark color of the asphalt pavement and the texture characteristics of the pavement, current asphalt pavement crack detection technology cannot meet the requirements of accuracy and efficiency. In this paper, we propose an end-to-end multi-scale full convolutional neural network to achieve the semantic segmentation of cracks in road images by learning the crack characteristics in the complex fine grain background of asphalt pavement. The method uses DenseNet and deconvolution network framework to achieve pixel-level detection and fuses features learned from different scales of convolutional kernels through a full convolutional network to obtain richer information on multi-scale features, allowing more detailed representation of crack features in high-resolution images. And the back end joins the SVM classifier to achieve crack classification after crack segmentation. Then we create a road test standard data set containing 12 cracks and evaluate it on the data. The experimental results show that the method achieves good segmentation effect for 12 types of cracks, and the crack segmentation for asphalt pavement is better than the most advanced methods. [ABSTRACT FROM AUTHOR]

Researchers used visual methods rigorously to improve brain tumor detection in MRI or CT scans, yet there remains a challenge to improve the detection accuracy. Further, the rise of deep learning methods improved tumor detection accuracy up to the mark. But again, many times, we face the challenges of having a bigger dataset and better computing power to achieve an improved and accurate trained model for every object classification problem. In this paper, we propose a deep learning framework single shot multi-box detector (SSD)-based model to detect tumors in the MRI scans. The proposed SSD model is the faster algorithm to detect the tumor even with the ability to detect the smallest spot in the low-resolution MRI scans. We additionally used a lightweight neural network architecture MobileNet v2 with SSD for faster and accurate object classification. The experimental results showed 98% accuracy with the proposed method after training with the smallest dataset of 250 MRI scans. We used the Kaggle database for training and testing the proposed model. [ABSTRACT FROM AUTHOR]

Influenced by national policies and macro-economic environment, large domestic enterprises is actively promoting strategic transformation to enhance their core competitiveness, and performance evaluation of enterprises' innovation capacity has become a hot topic in recent years. This paper proposes a performance evaluation method of enterprises' innovation capacity based on deep learning fuzzy system model and convolutional neural network analysis of innovation network. First of all, on account of the characteristics of breakthrough innovation and drawing on the traditional innovation performance evaluation model, this paper constructs a breakthrough innovation performance evaluation index system for enterprises from the six dimensions of main resource input, technology out-turn, process management, product performance, social value and commercial Value. Secondly, the introduction of machine learning of fuzzy convolutional neural network to assess the advancement execution of enterprises is of great significance for enterprise managers to find out the problems and causes of enterprises' innovation, optimize the allocation of enterprises' resources and further improve the innovation performance of enterprises. The experimental results show to verify the adequacy of the algorithm. [ABSTRACT FROM AUTHOR]

Deep learning (DL) algorithms, especially the convolutional neural network (CNN), have been proven as a newly developed tool in machinery intelligent diagnosis. However, the current CNN-based fault diagnosis studies usually consider features or images extracted from a single domain as model input. This single domain information may not reflect fault patterns comprehensively, leading to low modeling accuracy and inaccurate diagnostic results. To overcome this limitation, this paper proposes a new CNN-based fault diagnosis approach using image representation considering multi-domain features of vibration signals. First, multi-domain features of vibration signals are extracted. These extracted features are then used to construct a n × n matrix, and subsequently to form images by RGB color transformations. This image transformation technique allows for capturing complementary and rich diagnostic information from multiple domains. At last, these images associated with different mechanical defects are fed into a CNN model that is improved based on the classic LeNet-5 CNN architecture for fault diagnosis and identification. Comparative experiments with the traditional feature extraction methods as well as state-of-the-art CNN-based methods are also investigated. Experimental studies on rolling bearings validate the effectiveness and superiorities of the proposed approach. [ABSTRACT FROM AUTHOR]

Stroke is a major reason for disability and mortality in most of the developing nations. Early detection of stroke is highly significant in bio-medical research. Research illustrates that signs of stroke are reflected in the eye and may be analyzed from fundus images. A custom dataset of fundus images has been compiled for formulating an automated stroke detection algorithm. In this paper, a comparative study of hand-crafted texture features and convolutional neural network (CNN) has been recommended for stroke diagnosis. The custom CNN model has also been compared with five pre-trained models from ImageNet. Experimental results reveal that the recommended custom CNN model gives the best performance by achieving an accuracy of 95.8 %. [ABSTRACT FROM AUTHOR]

Neural decoding is a technology to analyze intentions produced by neural activities, which has important applications in military, medical, entertainment and so on. As a typical application, decoding electromyogram (EMG) signals into corresponding gestures is an important content. In order to improve the accuracy of EMG signals recognition, researchers often extract effective features from EMG signals and classify gestures by constructing a reasonable classifier. However, because of the stochasticity of the signals, this method is not robust enough. This paper proposes a convolutional neural network (CNN) based on feature fusion, which can automatically learn and classify features from time-domain(TD) and frequency-domain(FD). To make full use of information, two fusion methods are used and compared. Experiments show that the proposed fusion methods are superior to the traditional algorithm for both normal people and amputees, and have better performance compared with CNN method using only one kind of information. [ABSTRACT FROM AUTHOR]

The deployment of large-scale Convolutional Neural Networks (CNNs) in limited-power devices is hindered by their high computation cost and storage. In this paper, we propose a novel framework for CNNs to simultaneously achieve channel pruning and low-bit quantization by combining weight quantization with Sparse Group Lasso (SGL) regularization. We model this framework as a discretely constrained problem and solve it by Alternating Direction Method of Multipliers (ADMM). Different from previous approaches, the proposed method reduces not only model size but also computational operations. In experimental section, we evaluate the proposed framework on CIFAR datasets with several popular models such as VGG-7/16/19 and ResNet-18/34/50, which demonstrate that the proposed method can obtain low-bit networks and dramatically reduce redundant channels of the network with slight inference accuracy loss. Furthermore, we also visualize and analyze weight tensors, which showing the compact group-sparsity structure of them. [ABSTRACT FROM AUTHOR]

Machine learning techniques commonly used for intrusion detection systems (IDSs face challenges due to inappropriate features and class imbalance. A novel IDS comprises four stages: Pre-processing, Feature Extraction, Feature Selection, and Detection. Initial pre-processing balances input data using an improved technique. Features (statistical, entropy, correlation, information gain) are extracted, and optimal ones selected using Improved chi-square. Intrusion detection is performed by a hybrid model combining Bi-GRU and CNN classifiers, with optimized weight parameters using SI-BMO. The outputs from both classifiers are averaged for the result. The SI-BMO-based IDS is compared with conventional techniques Blue Monkey Optimization (BMO), Grasshopper Optimization Algorithm (GOA), Deer Hunting Optimization (DHO), Poor Rich Optimization (PRO), Long Short-Term Memory (LSTM), Support Vector Machine (SVM), Gated Recurrent Unit (GRU), Recurrent Neural Network (RNN) for performance evaluation. [ABSTRACT FROM AUTHOR]

Breast cancer is responsible for the deaths of hundreds of women every year. The manual identification of breast cancer has more difficulties, and have the possibility of error. Many imaging approaches are being researched for their potential to identify breast cancer (BC). Incorrect identification might sometimes result in unneeded therapy and diagnosis. Because of this, accurate identification of breast cancer may save a great number of patients from needing unneeded surgery and biopsies. Deep learning's (DL) performance in the processing of medical images has substantially increased as a result of recent breakthroughs in the sector. Because of their improved capacity to anticipate outcomes, deep learning algorithms are able to reliably detect BC from ultrasound pictures. Transfer learning is a kind of machine learning that reuses knowledge representations from public models that were built with the use of large-scale datasets. Transfer learning has been shown to often result in overfitting. The primary purpose of this research is to develop and provide suggestions for a deep learning model that is effective and reliable in the detection and classification of breast cancer. A tissue biopsy is obtained from the suspicious region in order to ascertain the nature of a breast tumor and whether or not it is cancerous. Tumors may take any of these forms. When the images have been reconstructed with the help of a variational autoencoder (VAE) and a denoising variational autoencoder (DVAE), a convolutional neural network (CNN) model is used. This will be the case because it opens up a new area of the field to be investigated. The histological subtypes of breast cancer are used in conjunction with the degree of differentiation to execute the task of breast cancer categorization. [ABSTRACT FROM AUTHOR]

A spectrum-image based representation of machine vibration signals with deep convolution neural network is proposed for machine fault classification in which the convolution layer is used for automatic feature extraction as an alternate to the conventional feature-based methods. Two different forms of spectrum representations are proposed, one based on the short time Fourier transform of the original signals and the other based on the short time Fourier transform of the intrinsic mode functions acquired by empirical mode decomposition. Empirical mode decomposition has its own merits in discriminating non stationary signals and the novelty of the work is to use the short time Fourier transform of intrinsic mode functions with deep convolution neural network model. The classification and validation accuracy of the model are investigated with respect to epochs. It is demonstrated that both spectrum-based techniques perform good with 100% model accuracies in a numerical experiment of binary classification on a bearing dataset that comprises of normal and faulty signals. In another experiment using milling data set, short time Fourier transform of intrinsic mode functions representation performs better with 100% training accuracy, F1 score of 0.8933 which is better than that of using short time Fourier transform of raw signals whose training accuracy is 64% and F1 score of 0.7486. The numerical study shows that the empirical mode decomposition based spectrum representation delivers the highest accuracy in the learning model obviating the necessity for independent feature extraction, feature selection, and dimension reduction. The numerical experiment is extended using empirical mode decomposition based spectrums for multiple class classification problems in bearing dataset. The confusion matrix obtained for 10 classes, shows that validation accuracy is 100% for all classes. The performance comparison throws light on the merits of empirical mode decomposition spectrum method over other state of the art methods. [ABSTRACT FROM AUTHOR]

COVID-19 (Coronavirus Disease of 2019) is one of the most challenging healthcare crises of the twenty-first century. The pandemic causes many negative impacts on all aspects of life and livelihoods. Although recent developments of relevant vaccines, such as Pfizer/BioNTech mRNA, AstraZeneca, or Moderna, the emergence of new virus mutations and their fast infection rate yet pose significant threats to public health. In this context, early detection of the disease is an important factor to reduce its effect and quickly control the spread of pandemic. Nevertheless, many countries still rely on methods that are either expensive and time-consuming (i.e., Reverse-transcription polymerase chain reaction) or uncomfortable and difficult for self-testing (i.e., Rapid Antigen Test Nasal). Recently, deep learning methods have been proposed as a potential solution for COVID-19 analysis. However, previous works usually focus on a single symptom, which can omit critical information for disease diagnosis. Therefore, in this study, we propose a multi-modal method to detect COVID-19 using cough sounds and self-reported symptoms. The proposed method consists of five neural networks to deal with different input features, including CNN-biLSTM for MFCC features, EfficientNetV2 for Mel spectrogram images, MLP for self-reported symptoms, C-YAMNet for cough detection, and RNNoise for noise-canceling. Experimental results demonstrated that our method outperformed the other state-of-the-art methods with a high AUC, accuracy, and F1-score of 98.6%, 96.9%, and 96.9% on the testing set. [ABSTRACT FROM AUTHOR]

Computer vision functions like object detection, image segmentation, and image classification were recently getting advance due to Convolutional Neural Networks (CNNs). In the food and agricultural industries, image classification plays a critical role in quality control. CNNs are made up of layers that alternate between convolutional, nonlinearity, and feature pooling. In this article, we proposed Fuzzy Pooling, a novel pooling approach that works based on fuzzy logic, that can increase the accuracy of the CNNs by replacing the conventional pooling layer. This proposed Fuzzy Pooling was put to the test with CIFAR-10 and SVHN data sets on single layer CNN, and it outperformed previous pooling strategies by achieving 92% and 97% classification accuracy. This proposed Fuzzy Pooling layer was replaced the Max Pooling layer in the ThinNet architecture, and it was trained using the back propagation method. It was demonstrated experimentally on the Lemon fruit data set to classify the fruits into three categories such as Good, Medium, and Poor. In order to classify the lemon fruit into three categories, the 1000 Fully Connected layer in ThinNet architecture was replaced with three Fully Connected layers. The Modified ThinNet architecture called ThinNet_FP was trained with a learning rate of 0.001 and achieved 97% accuracy in classifying the images and outperformed previous CNN architectures when trained on the same data set. [ABSTRACT FROM AUTHOR]

The face is a dominant biometric for recognizing a person. However, face recognition becomes challenging when there are severe changes in lighting conditions, i.e., illumination variations, which have been shown to have a more severe effect on recognition performance than the inherent differences between individuals. Most of the existing methods for tackling the problem of illumination variation assume that illumination lies in the large-scale component of a facial image; as such, the large-scale component is discarded, and features are extracted from small-scale components. Recently, it has been shown that large-scale component is also important; in addition, small-scale component contains detrimental noise features. Keeping this in view, we introduce a method for illumination invariant face recognition that exploits large-scale and small-scale components by discarding the illumination artifacts and detrimental noise using ContourletDS. After discarding the unwanted components, local and global features are extracted using a convolutional neural network (CNN) model; we examined three widely employed CNN models: VGG-16, GoogLeNet, and ResNet152. To reduce the dimensions of local and global features and fuse them, we employ linear discriminant analysis (LDA). Finally, ridge regression is used for recognition. The method was evaluated on three benchmark datasets; it achieved accuracies of 99.7%, 100%, and 79.76% on Extended Yale B, AR, and M-PIE, respectively. The comparison reveals that it outperforms the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Electroencephalogram (EEG) signals are essential in brain-computer interface systems. Nowadays, these signals are employed in various medical applications. In the past few years, EEG signals gain more attention in security systems to identify users, as these signals are unique for each individual. The current study explores deep learning frameworks for EEG-based user identification. Data from 107 users were considered for the study, which is acquired using 64 channels. Several experimental tests are performed over both convolutional neural network (CNN) and recurrent neural networks (RNN) using a 10-fold cross-validation process to check system effectiveness. In CNN, 1-D Convolutional layer is employed for the processing of EEG signals. In RNN, LSTM and GRU are used to check system accuracy. For performance measure various metrices were considered such as accuracy, precision, recall and kappa score. Acquired results suggest that gated recurrent unit (GRU) outperforms other models in terms of accuracy and complexity both. GRU model has 91.2% accuracy and has three layers only, which reduces the model's complexity. The training cost is also decreasing due to the low complexity of the model. [ABSTRACT FROM AUTHOR]