Search

The use of Artificial Intelligence (AI) to preserve and promote cultural heritage has experienced significant growth in recent years. Among the various areas of cultural heritage, numismatics have emerged as a particularly promising field where we can develop AI solutions. Numismatics refers to the study of coins, tokens, paper money, and medals, which play a critical role in understanding human history and culture. However, there are still limited resources available to help researchers and collectors in the identification of coins. This is due to the vast number of coins in circulation, which presents a significant challenge in developing smart tools for classification tasks. This paper aims to provide a contribution to this setting. In particular, we start by creating a new dataset called EURO-Coin, which consists of images showing the side of coins with reliefs and is designed to facilitate the training and testing of AI models for euro coin classification. Then, we propose two approaches that leverage Convolutional Neural Networks and self-attention layers to classify the country and value of the coins. In our experiments, we obtain an accuracy of 86.9% for country classification and an accuracy of 96.4% for value classification. Finally, we conduct an ablation study to evaluate the impact of the preprocessing activities and attention layers in our approaches. [ABSTRACT FROM AUTHOR]

The model integrates artificial intelligence (AI) and big data analytics, utilizing IoMT devices for data acquisition and Hadoop ecosystem for managing big data. The field of medical diagnosis is currently undergoing a remarkable transformation with the emergence of artificial intelligence (AI) techniques, particularly deep learning and big data analytics. By harnessing the power of deep learning and big data analytics, AI-based e-diagnosis has the potential to revolutionize healthcare delivery. [Extracted from the article]

Traditional sentiment analysis focuses on text-level sentiment mining, transforming sentiment mining into classification or regression problems, resulting in a sentiment analysis low accuracy rate. Sentiment analysis refers to the use of natural language processing, text analysis, and computational linguistics to systematically identify, extract, quantify, and study sentimental states. Therefore, more scholars have begun to focus on speech recognition and facial expression recognition research, and extracting and analysing people's sentiment tendencies can improve sentiment recognition accuracy. Traditional single-modal sentiment analysis can no longer meet people's needs. Therefore, this paper proposes a multimodal sentiment analysis method based on the multimodal sentiment analysis method that can obtain more sentimental information sources and help people make better decisions. The experimental results in this paper show that the highest recognition rates of CNN-SVM, RNN-SVM, and CRNN-SVM were 76.8%, 71.2%, and 93.5%, respectively. It can be seen that CRNN-SVM has the highest sentiment tendency recognition rate in deep learning, so it is suitable to apply CRNN-SVM to sentiment tendency analysis system design in this paper. The average accuracy rate of the system designed in this paper was 91%, and the stability was also very strong, which shows that the system designed in this paper is meaningful. The main contribution of this paper is based on the limitations of single-mode emotion analysis. It proposes a multimode emotion analysis method and introduces a convolutional neural network to help people obtain more emotional information sources to meet their needs. [ABSTRACT FROM AUTHOR]

In big data era, multi-source heterogeneous data become the biggest obstacle to data sharing due to its high dimension and inconsistent structure. Using text classification to solve the ontology construction and mapping problem of multi-source heterogeneous data can not only reduce manual operation, but also improve the accuracy and efficiency. This paper proposes an ontology construction and mapping scheme based on hybrid neural network and autoencoder. Firstly, the proposed text classification method uses the multi-core convolutional neural network to capture local features and uses the improved Bidirectional Long Short-Term Memory network to compensate for the shortcomings of the convolutional neural network that cannot obtain context-related information. Secondly, a similarity matching method is used for ontology mapping, which integrate autoencoder to improve anti-interference ability. We have carried out several sets of experiments to test the validity of the proposed ontology construction and mapping scheme. [ABSTRACT FROM AUTHOR]

For detecting and resolving the various types of malware, novel techniques are proposed, among which deep learning algorithms play a crucial role. Although there has been a lot of research on the development of DL-based mobile malware detection approaches, they were not reviewed in detail yet. This paper aims to identify, assess, and synthesize the reported articles related to the application of DL techniques for mobile malware detection. A Systematic Literature Review is performed in which we selected 40 journal articles for in-depth analysis. This SLR presents and categorizes these articles based on machine learning categories, data sources, DL algorithms, evaluation parameters & approaches, feature selection techniques, datasets, and DL implementation platforms. The study also highlights the challenges, proposed solutions, and future research directions on the use of DL in mobile malware detection. This study showed that Convolutional Neural Networks and Deep Neural Networks algorithms are the most used DL algorithms. API calls, Permissions, and System Calls are the most dominant features utilized. Keras and Tensorflow are the most popular platforms. Drebin and VirusShare are the most widely used datasets. Supervised learning and static features are the most preferred machine learning and data source categories. [ABSTRACT FROM AUTHOR]

Brain–computer interfacing (BCI) research, fueled by deep learning, integrates insights from diverse domains. A notable focus is on steady-state visual evoked potential (SSVEP) in BCI applications, requiring in-depth assessment through deep learning. EEG research frequently employs SSVEPs, which are regarded as normal brain responses to visual stimuli, particularly in investigations of visual perception and attention. This paper tries to give an in-depth analysis of the implications of deep learning for SSVEP-adapted BCI. A systematic search across four stable databases (Web of Science, PubMed, ScienceDirect, and IEEE) was developed to assemble a vast reservoir of relevant theoretical and scientific knowledge. A comprehensive search yielded 177 papers that appeared between 2010 and 2023. Thence a strict screening method from predetermined inclusion criteria finally generated 39 records. These selected works were the basis of the study, presenting alternate views, obstacles, limitations and interesting ideas. By providing a systematic presentation of the material, it has made a key scholarly contribution. It focuses on the technical aspects of SSVEP-based BCI, EEG technologies and complex applications of deep learning technology in these areas. The study delivers more penetrating reporting on the latest deep learning pattern recognition techniques than its predecessors, together with progress in data acquisition and recording means suitable for SSVEP-based BCI devices. Especially in the realms of deep learning technology orchestration, pattern recognition techniques, and EEG data collection, it has effectively closed four important research gaps. To increase the accessibility of this critical material, the results of the study take the form of easy-to-read tables just generated. Applying deep learning techniques in SSVEP-based BCI applications, as the research shows, also has its downsides. The study concludes that a radical framework will be presented which, includes intelligent decision-making tools for evaluation and benchmarking. Rather than just finding a comparable or similar analogy, this framework is intended to help guide future research and pragmatic applications, and to determine which SSVEP-based BCI applications have succeeded at responsibility for what they set out with. [ABSTRACT FROM AUTHOR]

Visual question answering is a challenging task that necessitates for sophisticated reasoning over the visual elements to provide an accurate answer to a question. Majority of the state-of-the-art VQA models are only applicable to English questions. However, applications such as visual assistance and tourism necessitate the incorporation of multilingual VQA systems. This paper presents an effective deep learning framework for Malayalam visual question answering (MVQA), which can answer a specific natural language question about an image in Malayalam. As there is no available dataset in English–Malayalam VQA, a MVQA dataset was created by translating English question–answer pairs from the visual genome dataset. The paper proposes an attention-driven MVQA model on the developed dataset. The proposed MVQA model uses a deep learning-based co-attention mechanism to jointly learn the attention for images and Malayalam questions. A second-order multimodal factorized high-order pooling is used for multi modal feature fusion. Different VQA models using combinations of classical CNNs and RNNs were experimented on the developed MVQA dataset, and the performance was compared against the proposed attention-driven model. Experimental results show that the proposed attention-driven MVQA model achieves state-of-the-art results as compared to other models for MVQA on the custom Malayalam VQA dataset. [ABSTRACT FROM AUTHOR]

The inspection and diagnosis of building engineering involve rail surface defect detection, which plays a crucial role in assessing the quality of railway tracks. However, achieving accurate detection remains a significant challenge for infrastructure construction due to the challenging factors, such as complex background, poor texture, the irregular shapes and sizes of rail surface defects, etc. To address these challenges, combined with vision sensor, this paper proposes an automatic vision-based rail surface defect detection network, which leverages a combination of convolutional neural network (CNN) and transformer. Specifically, an effective encoding path based on an improved Res2N-et is presented to obtain stronger contextual information. Meanwhile, to address the limitation of feature representation on global context information, the transformer block is introduced into the bottleneck layer to capture essential global context information. Moreover, an attention-based edge enhancement block is proposed to mitigate the loss of local edge details and strengthen the focus on key feature information, to acquire more discriminant features. To further enrich the feature representation capability, this paper introduces an efficient feature aggregation block, which combines the global features extracted by the transformer and the local features extracted by CNN, to achieve effective feature complementarity and enhances overall detection performance. Combined with public NRSD-MN dataset, proposed model obtains 87.3 % and 77.9 % on PA metric, 85.9 % and 83.4 % in mIoU metric among artificial dataset with natural dataset. Meanwhile, it also obtains 84.0 % PA metric and 87.0 % mIoU metric in SEPR dataset. Experiments have proven the superiority of proposed model though the performance comparison with latest models, which presents an effective and promising rail surface defect detection solution for railway track quality assessment, and can be instrumental in ensuring safe and efficient railway operations. [ABSTRACT FROM AUTHOR]

Mental stress is a common problem that affects individuals all over the world. Stress reduces human functionality during routine work and may lead to severe health defects. Early detection of stress is important for preventing diseases and other negative health-related consequences of stress. Several neuroimaging techniques have been utilized to assess mental stress, however, due to its ease of use, robustness, and non-invasiveness, electroencephalography (EEG) is commonly used. This paper aims to fill a knowledge gap by reviewing the different EEG-related deep learning algorithms with a focus on Convolutional Neural Networks (CNNs) and Long Short-Term Memory networks (LSTMs) for the evaluation of mental stress. The review focuses on data representation, individual deep neural network model architectures, hybrid models, and results amongst others. The contributions of the paper address important issues such as data representation and model architectures. Out of all reviewed papers, 67% used CNN, 9% LSTM, and 24% hybrid models. Based on the reviewed literature, we found that dataset size and different representations contributed to the performance of the proposed networks. Raw EEG data produced classification accuracy around 62% while using spectral and topographical representation produced up to 88%. Nevertheless, the roles of generalizability across different deep learning models and individual differences remain key areas of inquiry. The review encourages the exploration of innovative avenues, such as EEG data image representations concurrently with graph convolutional neural networks (GCN), to mitigate the impact of inter-subject variability. This novel approach not only allows us to harmonize structural nuances within the data but also facilitates the integration of temporal dynamics, thereby enabling a more comprehensive assessment of mental stress levels. [ABSTRACT FROM AUTHOR]

In recent years, transformers, initially developed for language, have been successfully applied to visual tasks. Vision transformers have been shown to push the state of the art in a wide range of tasks, including image classification, object detection, and semantic segmentation. While ample research has shown promising results in art attribution and art authentication tasks using convolutional neural networks, this paper examines whether the superiority of vision transformers extends to art authentication, improving, thus, the reliability of computer-based authentication of artworks. Using a carefully compiled dataset of authentic paintings by Vincent van Gogh and two contrast datasets, we compare the art authentication performances of Swin transformers with those of EfficientNet. Using a standard contrast set containing imitations and proxies (works by painters with styles closely related to van Gogh), we find that EfficientNet achieves the best performance overall. With a contrast set that only consists of imitations, we find the Swin transformer to be superior to EfficientNet by achieving an authentication accuracy of over 85%. These results lead us to conclude that vision transformers represent a strong and promising contender in art authentication, particularly in enhancing the computer-based ability to detect artistic imitations. [ABSTRACT FROM AUTHOR]

The automatic reading of pointer meters is significantly important to data monitoring and efficient measurement in the industrial field. However, the existing automatic reading method can not obtain accurate performance in natural scenarios and present no satisfactory application effects in industrial fields (such as power stations and gas stations). In this paper, a novel automatic reading method for pointer meters based on deep learning is proposed, which contains five stages. Stage-1: the object detection algorithm Yolov4 and the feature optimization module IFF are used to locate the target meter. Stage-2: Semantic segmentation model is applied to extract the pointer area based on Anam-Net. Stage-3: the character detection algorithm CRAFT and the text recognition algorithm E2E-MLT are combined and used to recognize the scale text and unit on the meter. Stage-4: the scale area of the meter is converted to the polar coordinate system, and a lightweight convolutional neural network is designed to locate the main scale line. And finally in Stage-5: the reading data are calculated according to the outputs of the above-mentioned deep learning models. The experiment results show that the reading method proposed in this paper has higher accuracy and robustness than those of the existing approaches and obtains satisfactory application effects in the industrial field. [ABSTRACT FROM AUTHOR]

Image classification can effectively manage and organize images, laying a good foundation for the work in multiple fields of image processing. With the rise of Internet technology and social networks, the number of digital images has increased dramatically and there are more and more applications. People also use intuitive pictures instead of words when expressing emotions and information. A large number of digital images need to be managed, analyzed, and retrieved. This urgently requires more efficient and accurate image classification technology. Deep learning is a learning method that extends traditional neural networks, simulating the process of gradual abstraction of human brain cognition. The number of hidden layers is deepened, and features can be learned automatically. This paper has completed the traditional image noise detection and segmentation experiment based on convolutional neural network. We introduced the various parts of the convolutional neural network, including the convolutional layer, activation function, fully connected layer, etc. Noise detection is completed based on Fast RCNN on the MSTAR data set containing the background. The model fully combines the advantages of the isotropic model, the Perona-Malik model, and the second-order directional derivative. In order to better maintain the edge structure information of the image, the structure information is extracted based on the original noise image, and the improved ID model and the improved PM model are directionally diffused along the edge tangent direction of the original noise image. Considering the local structure information of the image, we use the slice similarity modulus value as the edge detector of the improved PM model, and use the slice similarity modulus value to construct a new weighting function to adaptively balance the relative weights of the improved ID model and the improved PM model. Simulation and measured data verify the effectiveness of this network in removing image coherent speckle noise. We compare and analyze it with existing denoising methods. The use of visual evaluation and objective evaluation indicators to evaluate the denoising effect and calculation efficiency shows the advantages of the network in this paper in terms of denoising effect, calculation time and space complexity. [ABSTRACT FROM AUTHOR]

In cloud services and Internet-of-Things (IoT) applications, the cybersecurity in video transmission technologies has drawn much attention in recent researches due to the rapid growth of cyber-risks on both individuals and institutions. Unfortunately, the spoofing attack, a kind of cyber-risk, has increased the number of cyber-criminals in data transfer applications without being detected, especially in smart cities. Several applications based on online video communications, such as online testing and video conferences, are involved in smart cities. The video displays various variations of a person, which makes face recognition an important concept in security implementation. The face spoofing attacks are mainly based on the person's face replication by replaying a video or by printed photos. Therefore, video forgery detection and related spoof attack detection have become a new topic in cybersecurity research. From this perspective, this paper presents a deep learning approach for video face forensic detection with a cyber facial spoofing attack using two methodologies. The first methodology is based on a convolutional neural network (CNN) to extract features from the input video frames. The model has five convolutional layers followed by five pooling layers. The second methodology is based on convolutional long short-term memory (ConvLSTM). This model comprises two pooling layers, two convolutional layers, and a convolutional LSTM layer. Each methodology includes a fully-connected layer to interconnect between the feature map resulting from the feature extraction process and the classification layer. A SoftMax layer performs the classification task in each method. This paper aims to achieve an optimum modality for face forensic detection to overcome spoofing attacks. Simulation results reveal that the ConvLSTM with CNN methodology achieves better classification results as the extracted features are more comprehensive than those of other conventional approaches. Also, it achieves an accuracy of 99%, and the works presented in the literature achieve an accuracy levels up to 95%. [ABSTRACT FROM AUTHOR]

Natural disasters can be unpredictable and catastrophic. Even after the event, the repercussions are prolonged due to the incompetence of disaster management strategies. To mitigate the effects of a natural hazard, disaster management teams have to rapidly come forth with innovative plans of action. To contain the damage that a disaster causes, the response time, and preparedness of a disaster management team is crucial. Having a forewarning about the nature of the disaster can prove to be beneficial for the management team. Remote/unreachable areas such as deep parts of the forests, far-flung rural areas, oceans and other hard to reach locations are at a higher risk of receiving poorer aid and response due to the lack of communication and connectivity with the rest of the world. Our paper provides a solution for this particular problem by suggesting UAV/Drones for surveillance and monitoring in these inaccessible, disaster struck areas. The drones not only monitor the situation but constantly click pictures and send them to a base station, where the images are used to collect insights about the type of disaster that has to be dealt with. The paper proposes a deep learning model based on feature concatenation for classification of disasters which can be deployed at the base station, and can receive data in the form of images from a UAV/drone hovering over the affected place. The proposed model is efficient and able to achieve a higher accuracy as compared to the leading CNN models and closely related recent works as well. [ABSTRACT FROM AUTHOR]

IoT devices convert billions of objects into data-generating entities, enabling them to report status and interact with their surroundings. This data comes in various formats, like structured, semi-structured, or unstructured. In addition, it can be collected in batches or in real time. The problem now is how to benefit from all of this data gathered by sensing and monitoring changes like temperature, light, and position. In this paper, we propose a predictive analytics framework constructed on top of open-source technologies such as Apache Spark and Kafka. The framework focuses on forecasting temperature time series data using traditional and deep learning predictive analytics methods. The analysis and prediction tasks were performed using Autoregressive Integrated Moving Average (ARIMA), Seasonal Autoregressive Integrated Moving Average (SARIMA), Long Short-Term Memory (LSTM), and a novel hybrid model based on Convolution Neural Network (CNN) and LSTM. The purpose of this paper is to determine whether and how recently developed deep learning-based models outperform traditional algorithms in the prediction of time series data. The empirical studies conducted and reported in this paper demonstrate that deep learning-based models, specifically LSTM and CNN-LSTM, exhibit superior performance compared to traditional-based algorithms, ARIMA and SARIMA. More specifically, the average reduction in error rates obtained by LSTM and CNN-LSTM models were substantial when compared to other models indicating the superiority of deep learning. Moreover, the CNN-LSTM-based deep learning model exhibits a higher degree of closeness to the actual values when compared to the LSTM-based model. [ABSTRACT FROM AUTHOR]

In an iron ore pelletization plant, pellets are produced inside a rotating disc pelletizer. Online pellet size distribution is an important performance indicator of the pelletization process. Image processing-based system is an effective solution for online size analysis of iron ore pellets. This paper proposes a machine learning algorithm for estimating the size class of the pellets during their production by imaging from an area inside the disc pelletizer. Instead of computing the size of each individual pellets in the acquired image, this method proposes a qualitative approach to get the overall size estimate of the pellets in production. The key idea of this paper is to find out whether the disc is producing VERY SMALL, SMALL, MEDIUM, or BIG-sized pellets. A weighted average ensemble of different convolutional neural networks such as VGG16, Mobilenet, and Resnet50 is used to achieve this objective. Furthermore, batch normalization is applied to improve the estimation performance of the proposed model. A novel data augmentation method is applied to the in situ captured images to create the data set used to train and evaluate the proposed ensemble of CNN models. Results of experiments indicate that it is possible to detect the operating state of the pelletization disc by acquiring images from the inside area of the disc with sufficient accuracy. [ABSTRACT FROM AUTHOR]

This paper aims to propose a new technique for identifying and categorizing malevolent Internet traffic within the context of security for smart devices. Given the rising usage of smart devices, including mobile phones, wearables, smart transportation-based devices, and the Internet of Things, concerns regarding their security are increasing. The need to develop effective security measures arises from the potential for attackers to compromise user data. In this study, we introduce an innovative approach that combines deep learning techniques, specifically convolutional neural networks (CNN), with long short-term memory (LSTM) for the purpose of detecting and categorizing malevolent Internet traffic. The objective of the proposed technique is to address the challenges related to time estimation by focusing on level prediction, resulting in a substantial reduction in prediction time for the identification of malevolent traffic. We utilize bidirectional long short-term memory–CNN (BI-LSTM–CNN) to identify malevolent communication and provide support for voice input. Experimental outcomes illustrate the effectiveness of our proposed technique in terms of precision, accuracy, F1 factor, false acceptance rate (FAR), false positive rate (FPR), and detection rate. In comparison with existing methods for detecting malevolent traffic, our approach achieves a 99.62% traffic detection rate, 99.98% accuracy, and 0.01% FAR, whereas the accuracy, detection rate of malevolent traffic, and FAR of existing methods are 99.88%, 97.32%, and 4.31%, respectively. These outcomes emphasize the superior performance and analysis of our technique, rendering it a valuable contribution to the realm of smart device security. In summary, this paper proposes a novel BI-LSTM–CNN technique for detecting malevolent traffic in smart devices. The proposed methodology tackles time estimation challenges and exhibits superior performance when compared to existing techniques. [ABSTRACT FROM AUTHOR]

The usage of intelligent IoT devices is exponentially rising, and so the possibility of attacks in the IoT surfaces. The deep leaning algorithms are competent for directing the sanctuary investigation of IoT systems but have not upgraded the analysis of potential attacks in IoT. This paper aims to advance deep learning methods to create upgraded security strategies for IoT frameworks quickly. The study of the IoT security threats identified with inalienable or recently presented risks is done. Also, this paper does a quick examination of different possible attack surfaces for the IoT framework, and the potential risks identified with each character. The systematic survey of deep learning methods for IoT security and the existence of the chances, focal points, and weaknesses of every strategy opens the door significant for future research. [ABSTRACT FROM AUTHOR]

Hyperspectral imaging is a highly advanced and sophisticated method for capturing images in hundreds of narrow, contiguous spectral bands. However, processing and analyzing such large amounts of data are challenging. Deep learning algorithms, especially those based on convolutional neural networks (CNNs), effectively extract rich feature representations from complex datasets, such as hyperspectral images (HSIs). These representations capture high-level patterns and characteristics that can facilitate accurate classification. The success of this approach has led to its widespread use in various remote sensing applications. However, the standard CNN inputs and outputs are scalars, ignoring the relative position relationships between features. In this paper, we propose a hierarchical capsule network for HSI classification. This network incorporates a multi-level convolutional structure for feature extraction and fusion. It utilizes convolutional feature maps of various depths to generate initial capsules, followed by vector computation using capsule neurons and a weight matrix to encode spatial location relationships among features. Furthermore, the shallow convolution of the hierarchical capsule network is pre-trained based on transfer learning to further improve the performance of HSI classification. According to experimental results, the proposed method for hyperspectral image classification has been found to outperform other state-of-the-art deep learning models on four benchmark datasets. [ABSTRACT FROM AUTHOR]

Acute lymphocytic leukemia (ALL) is a common serious cancer in white blood cells (WBC) that advances quickly and produces abnormal cells in the bone marrow. Cancerous cells associated with ALL lead to impairment of body systems. Microscopic examination of ALL in a blood sample is applied manually by hematologists with many defects. Computer-aided leukemia image detection is used to avoid human visual recognition and to provide a more accurate diagnosis. This paper employs the ensemble strategy to detect ALL cells versus normal WBCs using three stages automatically. Firstly, image pre-processing is applied to handle the unbalanced database through the oversampling process. Secondly, deep spatial features are generated using a convolution neural network (CNN). At the same time, the gated recurrent unit (GRU)-bidirectional long short-term memory (BiLSTM) architecture is utilized to extract long-distance dependent information features or temporal features to obtain active feature learning. Thirdly, a softmax function and the multiclass support vector machine (MSVM) classifier are used for the classification mission. The proposed strategy has the resilience to classify the C-NMC 2019 database into two categories by using splitting the entire dataset into 90% as training and 10% as testing datasets. The main motivation of this paper is the novelty of the proposed framework for the purposeful and accurate diagnosis of ALL images. The proposed CNN-GRU-BiLSTM-MSVM is simply stacked by existing tools. However, the empirical results on C-NMC 2019 database show that the proposed framework is useful to the ALL image recognition problem compared to previous works. The DenseNet-201 model yielded an F1-score of 96.23% and an accuracy of 96.29% using the MSVM classifier in the test dataset. The findings exhibited that the proposed strategy can be employed as a complementary diagnostic tool for ALL cells. Further, this proposed strategy will encourage researchers to augment the rare database, such as blood microscopic images by creating powerful applications in terms of combining machine learning with deep learning algorithms. [ABSTRACT FROM AUTHOR]

Patients with deaths from COVID-19 often have co-morbid cardiovascular disease. Real-time cardiovascular disease monitoring based on wearable medical devices may effectively reduce COVID-19 mortality rates. However, due to technical limitations, there are three main issues. First, the traditional wireless communication technology for wearable medical devices is difficult to satisfy the real-time requirements fully. Second, current monitoring platforms lack efficient streaming data processing mechanisms to cope with the large amount of cardiovascular data generated in real time. Third, the diagnosis of the monitoring platform is usually manual, which is challenging to ensure that enough doctors online to provide a timely, efficient, and accurate diagnosis. To address these issues, this paper proposes a 5G-enabled real-time cardiovascular monitoring system for COVID-19 patients using deep learning. Firstly, we employ 5G to send and receive data from wearable medical devices. Secondly, Flink streaming data processing framework is applied to access electrocardiogram data. Finally, we use convolutional neural networks and long short-term memory networks model to obtain automatically predict the COVID-19 patient's cardiovascular health. Theoretical analysis and experimental results show that our proposal can well solve the above issues and improve the prediction accuracy of cardiovascular disease to 99.29%. [ABSTRACT FROM AUTHOR]

Experiments on the CheXpert dataset, which is a publicly available multi-disease chest radiograph dataset, and the TBX11K dataset show that the proposed model generates identical results. The MOFPA-WT is tested using a standard EEG signal processing dataset, namely the EEG motor movement/imagery dataset. [Extracted from the article]

Microscopic image analysis technology helps solve the inadvertences of artificial traditional methods in disease, wastewater treatment, and environmental change monitoring analysis. Convolutional neural network (CNN) play an important role in microscopic image analysis. Image segmentation, in which U-Net is increasingly applied in microscopic image segmentation, is a crucial step in detection, tracking, monitoring, feature extraction, modelling, and analysis. This paper comprehensively reviews the development history of U-Net, analyses several research results of various segmentation methods since the emergence of U-Net, and conducts a comprehensive review of related papers. This paper summarised the improved methods of U-Net and then listed the existing significance of image segmentation techniques and their improvements introduced over the years. Finally, focusing on the different improvement strategies of U-Net in different papers, the related work of each application target is reviewed according to detailed technical categories to facilitate future research. Researchers can see the dynamics of the transmission of technological development and keep up with future trends in this interdisciplinary field. [ABSTRACT FROM AUTHOR]

Dry eye disease (DED) is a chronic eye disease and a common complication among the world's population. Evaporation of moisture from tear film or a decrease in tear production leads to an unstable tear film which causes DED. The tear film breakup time (TBUT) test is a common clinical test used to diagnose DED. In this test, DED is diagnosed by measuring the time at which the first breakup pattern appears on the tear film. TBUT test is subjective, labour-intensive and time-consuming. These weaknesses make a computer-aided diagnosis of DED highly desirable. The existing computer-aided DED detection techniques use expensive instruments for image acquisition which may not be available in all eye clinics. Moreover, among these techniques, TBUT-based DED detection techniques are limited to finding only tear film breakup area/time and do not identify the severity of DED, which can essentially be helpful to ophthalmologists in prescribing the right treatment. Additionally, a few challenges in developing a DED detection approach are less illuminated video, constant blinking of eyes in the videos, blurred video, and lack of public datasets. This paper presents a novel TBUT-based DED detection approach that detects the presence/absence of DED from TBUT video. In addition, the proposed approach accurately identifies the severity level of DED and further categorizes it as normal, moderate or severe based on the TBUT. The proposed approach exhibits high performance in classifying TBUT frames, detecting DED, and severity grading of TBUT video with an accuracy of 83%. Also, the correlation computed between the proposed approach and the Ophthalmologist's opinion is 90%, which reflects the noteworthy contribution of our proposed approach. [ABSTRACT FROM AUTHOR]

Deep convolutional neural networks (DCNNs) are one of the most advanced techniques for classifying images in a range of applications. One of the most prevalent cancers that cause death in women is breast cancer. For survival rates to increase, early detection and treatment of breast cancer is essential. Deep learning (DL) can help radiologists diagnose and classify breast cancer lesions. This paper proposes a computer-aided system based on DL techniques for automatically classify breast cancer tumors in histopathological images. There are nine DCNN architectures used in this work. Four schemes are performed in the proposed framework to find the best approach. The first scheme consists of pre-trained DCNNs based on the transfer learning concept. The second scheme performs feature extraction of the DCNN architectures and uses a support vector machine (SVM) classifier for evaluation. The third one performs feature integration to show how the integrated deep features may enhance the SVM classifiers' accuracy. Finally, in the fourth scheme, the Chi-square (χ2) feature selection method is applied to reduce the large feature size in the feature integration step. The results of the proposed system present a promising performance for breast cancer classification with an accuracy of 99.24%. The system performance shows that the proposed tool is suitable to assist radiologists in diagnosing breast cancer tumors. [ABSTRACT FROM AUTHOR]

Haphazard urbanization has resulted in development of unplanned informal settlements, also called slums. Citizens residing in these dense settlements are generally deprived of essential sustainable drivers, i.e., green cover and open spaces (GOS). Such limitations can have an adverse impact on various aspects such as physiological, economical, health, and quality of life. Much research has been done on applying Machine Learning (ML) techniques for identifying slums at the city scale. However, no study explicitly addressed the problem of GOS extraction inside the informal settlements. This paper aims to fill this gap by training three modified Convolution Neural Network (CNN) models, i.e., VGG16-UNet, MobileNetV2-UNet, DeepLabV3 + on manually labeled high-resolution satellite imagery for the city of Mumbai. All three models performed excellently and achieved the state-of-the art overall accuracy of around 95%, for the considered classes. Out of the three models, VGG16-UNet performed the best and hence was applied to the whole city for developing green, open, and green and open indices. The indices were further used to establish a quantitative relation between various built-up typologies and GOS. It was found that informal settlements have significantly less green and open space than other built-up typologies. Stark GOS difference between slums and its neighborhood planned residential was observed, and hence, requires urgent attention of the city managers. The outcome presented in the paper can be used as input to various decision-making systems concerning better informal space planning to better plan the GOS, which eventually can lead to development of sustainable cities. [ABSTRACT FROM AUTHOR]

In recent years, image perception such as enhancement, classification and object detection with deep learning has achieved significant successes. However, in real world under extreme conditions, the training of a deep learning model often yields low accuracy, low efficiency in feature extraction and generalizability, due to the inner uncourteous and uninterpretable characteristics. In this paper, a maximal-entropy-attention-based convolutional neural network (MEA-CNN) framework is proposed. A maximum entropy algorithm is first used for image feature pre-extraction. An attention mechanism is then proposed by combining the extracted features on original images. By applying the mechanism, the key areas of an image are enhanced, and noised area can be ignored. Afterward, the processed images are transferred into region convolutional neural network, which is a well-known pre-trained CNN model, for further feature learning and extraction. Finally, two real-world experiments on traffic sign recognition and road surface condition monitoring are designed. The results show that the proposed framework has high testing accuracy, with improvements of 17% and 2.9%, compared with some other existing methods. In addition, the features extracted by the model are more easily interpretable. [ABSTRACT FROM AUTHOR]

In India, rice crops are very significant. Rice cultivation comprises several phases, and it is crucial to keep an eye on the crop's development to avoid any leaf diseases and to provide a good yield. To avoid yield loss, crop diseases need to be determined at the initial stage. Deep learning-based pre-trained CNN architecture is used in this study to identify rice leaf diseases. This paper discusses four different CNN architectures to classify and identify healthy and diseased leaves such as Brown spot, Hispa, and Leaf Blast. Initially, to avoid vanishing gradient problems that degrade the performance of the Network, ResNet34 and ResNet50 are used. Even though the CNN model performs the feature extraction, Self-attention with ResNet18 and ResNet34 architecture is utilized to improve the feature selection process. As a result of enhanced feature extraction, the accuracy of rice leaf disease identification and classification has improved. Finally, high accuracy of 98.54% is achieved with the proposed ResNet34 with self-attention architecture when compared to other CNN models used in this paper. In terms of multiclass classification, the proposed model offers improved outcomes when compared to state-of-the-art techniques. [ABSTRACT FROM AUTHOR]

Manual data annotation is a time consuming activity. A novel strategy for automatic training of the CAPTCHA breaking system with no manual dataset creation is presented in this paper. We demonstrate the feasibility of the attack against a text-based CAPTCHA scheme utilizing similar network infrastructure used for Denial of Service attacks. The main goal of our research is to present a possible vulnerability in CAPTCHA systems when combining the brute-force attack with transfer learning. The classification step utilizes a simple convolutional neural network with 15 layers. Training stage uses automatically prepared dataset created without any human intervention and transfer learning for fine-tuning the deep neural network classifier. The designed system for breaking text-based CAPTCHAs achieved 80% classification accuracy after 6 fine-tuning steps for a 5 digit text-based CAPTCHA system. The results presented in this paper suggest, that even the simple attack with a large number of attacking computers can be an effective alternative to current CAPTCHA breaking systems. [ABSTRACT FROM AUTHOR]

Nowadays, the volume of online information is growing and it is difficult to find the required information. Effective strategies such as recommender systems are required to overcome information overload. Collaborative filtering is a widely used type of recommender system in e-commerce environments and can simply provide suggestions for users. Recently, deep learning approaches were applied in collaborative filtering to tackle some drawbacks. This systematic review aims to provide a comprehensive review of recent research efforts on deep learning-based collaborative filtering recommender systems. We explain the research methodology and paper selection process and the search query. 102 papers are selected out of 719 papers that were published between 2019 and May 2023. Furthermore, the approaches in the selected papers are classified into two main categories: memory-based and model-based techniques. The main ideas, advantages, disadvantages, used tools, type of neural network, applications, and evaluation parameters of each selected paper are also discussed in detail. It was found that CNN (Convolutional Neural Network), AE (Autoencoder), DNN (Deep neural network), and Hybrid networks are the four mostly used neural networks in recommender systems. Also, Python, MATLAB, and Java are the most frequently used tools in the reviewed papers. Regarding the applications of the recommender systems in the reviewed papers, movies, products, and music recommendation are three most frequent applications. We point out the open issues and future research directions. Some key challenges such as cold start, data sparsity, scalability, and accuracy are still open to be addressed. [ABSTRACT FROM AUTHOR]

Combining convolutional neural networks (CNNs) and recurrent neural networks (RNNs) produces a powerful architecture for video classification problems as spatial–temporal information can be processed simultaneously and effectively. Using transfer learning, this paper presents a comparative study to investigate how temporal information can be utilized to improve the performance of video classification when CNNs and RNNs are combined in various architectures. To enhance the performance of the identified architecture for effective combination of CNN and RNN, a novel action template-based keyframe extraction method is proposed by identifying the informative region of each frame and selecting keyframes based on the similarity between those regions. Extensive experiments on KTH and UCF-101 datasets with ConvLSTM-based video classifiers have been conducted. Experimental results are evaluated using one-way analysis of variance, which reveals the effectiveness of the proposed keyframe extraction method in the sense that it can significantly improve video classification accuracy. [ABSTRACT FROM AUTHOR]

In this paper, offer a new framework for skin disease image recognition using deep learning techniques and local descriptor encoding approaches. For the purpose of detecting melanoma early, skin lesions must be accurately classified. In this research, an automatic image preprocessing approach is proposed for the removal of noise artefacts in photographs, including thin and thick hair objects, surgical ink markings, dark halo effects, and ebony frames. Due to hazy contrasts and distortions at the border margins, segmenting images are quite challenging. So, this research suggests a partitioning technique based on a fuzzy gray-level co-occurrence matrix (GLCM) that is both effective and adaptive. An alternative to convolutional neural networks (CNN) is proposed: the capsule-based network. An object's existence and the relationship between its functions are represented by a group of neurons (in logical units) that make up a vector called a capsule. While synthetic product neural networks use max-pooling layers to define capsule coupling between subsequent layers, capsule networks repeatedly utilise a dynamic routing technique to do so. Alternatively said, the routing-by-agreement approach offers learning between capsule layers. To assess the efficacy of the F-CapsNet technique, three widely used datasets—the ISIC 2017 Challenge, the 2019 Challenge, and the PH2 datasets—are employed. The suggested technique has an average accuracy of 99.16% for the ISBI 2017 test dataset and 99.45% accuracy for the ISBI 2019 test dataset. Additionally, the PH2 test dataset shows that the suggested approach has an average accuracy of 98.42%. [ABSTRACT FROM AUTHOR]

Automatic text recognition system might serve as an important factor in creating a paperless environment through digitizing and processing the existing paper documents in the upcoming days. Handwritten recognition using deep learning methods has been widely explored by many researchers. The existence of large quantity of data and a variety of algorithmic innovations enable the ease of training deep neural networks. Different techniques have been initiated in the literature for recognizing text from handwritten documents. This paper proposes a hybrid handwritten text recognition (H2TR) model using deep neural networks that use the sequence-to-sequence (Seq2Seq) approach. This hybrid model makes use of the salient features of convolution neural network (CNN) and recurrent neural network (RNN) with long–short-term memory network (LSTM). It uses CNN to extract the features from the handwritten image. The features that are extracted are later modelled with a sequence-to-sequence approach and fed to RNN–LSTM for encoding the visual features and decoding the sequence of letters that are available in the handwritten image. The proposed model is tested with IAM and RIMES handwritten databases, which shows competitive letter accuracy and word accuracy results. [ABSTRACT FROM AUTHOR]

In recent years, with the change of lifestyle in Europe and America, the incidence of breast cancer in Chinese women is increasing. In order to find the model of breast cancer image screening and diagnosis with higher accuracy and better classification performance, this paper mainly constructs the breast cancer CT image detection model and the breast cancer screening model based on the convolution and deconvolution neural network (CDNN) through the convolution neural network (CNN). In this paper, the fuzzy C-means clustering algorithm (FCM) is used to improve and optimize the image of breast cancer, and the experimental results are analyzed. The optimized kernel fuzzy C-means clustering algorithm was tested on a common dataset to segment the region of interest more accurately. Our experiments show that the new deep learning model of this paper improves the automatic classification performance of breast cancer. In this paper, the research results of deep learning are applied to the medical field, and a new method based on CNN model for breast cancer screening and diagnosis is proposed, which provides a new idea for improving the artificial intelligence assisted medical diagnosis method. [ABSTRACT FROM AUTHOR]

With the leaps and bounds of computer performance and the advent of the era of big data, deep learning has drawn more and more attention from all walks of life. It can combine low-level features to form more abstract high-level features and describe the data more essentially. Therefore, it is widely used in various fields such as computer vision. Image segmentation is one of the most basic research topics in the field of computer vision. The main purpose is to extract regions of interest from images for later image processing research. In 3D reconstruction based on sequence images, the segmentation accuracy and speed of sequence images determine the quality and efficiency of target reconstruction. Therefore, when facing large-scale sequence images, the biggest problem is how to improve the segmentation speed while ensuring accuracy. Based on the above background, the research content of this article is an indoor scene segmentation algorithm based on full convolutional neural network. According to the characteristics of indoor application scenes, this paper proposes a fast convolutional neural network image segmentation method to segment the indoor scene image and construct the fast fully convolutional networks (FFCN) for indoor scene image segmentation uses inter-layer fusion to reduce the amount of network calculation parameters and avoid the loss of picture feature information by continuous convolution. In order to verify the effectiveness of the network, in this paper, a basic living object data set (XAUT data set) in an indoor environment is created. The XAUT data set is used to train the FFCN network under the Caffe framework to obtain an indoor scene segmentation model. In order to compare the effectiveness of the model, the structure of the worn FCN8s, FCN16s, and FCN32s models was fine-tuned, and the corresponding algorithm model for indoor scene segmentation was obtained by training with the XAUT data set. The experimental results show that the pixel recognition accuracy of all types of networks has reached 86%, and the mean IU ratio has reached more than 63%. The mean IU of the FCN8s network is the highest at 70.38%, but its segmentation speed is only 1/5 of FFCN. On the premise that other types of indicators are not much different, the average segmentation speed on FFCN fast segmentation convolutional neural network reaches 40 fps. It can be seen that the scale fusion technology can well avoid the loss of image feature information in the network convolution and reddening process. Compared with other FCN networks, it has a faster speed and is conducive to real-time image preprocessing. [ABSTRACT FROM AUTHOR]

The monitoring of physical activities and recognition of motion disorders belong to important diagnostical tools in neurology and rehabilitation. The goal of the present paper is in the contribution to this topic by (1) analysis of accelerometric signals recorded by wearable sensors located at specific body positions and by (2) implementation of deep learning methods to classify signal features. This paper uses the general methodology to analysis of accelerometric signals acquired during cycling at different routes followed by the global positioning system. The experimental dataset includes 850 observations that were recorded by a mobile device in the spine area (L3 verterbra) for cycling routes with the different slope. The proposed methodology includes the use of deep learning convolutional neural networks with five layers applied to signal values transformed into the frequency domain without specification of any signal features. The accuracy of discrimination between different motion patterns for the uphill and downhill cycling and recognition of 4 classes associated with different route slopes was 96.6% with the loss criterion of 0.275 for sigmoidal activation functions. These results were compared with those evaluated for selected sets of features estimated for each observation and classified by the support vector machine, Bayesian methods, and the two-layer neural network. The best cross-validation error of 0.361 was achieved for the two-layer neural network model with the sigmoidal and softmax transfer functions. Our methodology suggests that deep learning neural networks are efficient in the assessment of motion activities for automated data processing and have a wide range of applications, including rehabilitation, early diagnosis of neurological problems, and possible use in engineering as well. [ABSTRACT FROM AUTHOR]

Retinal fundus image classification is pivotal for the early detection of various eye disorders. Leveraging advancements in deep learning, this research introduces a robust framework for retinal fundus image classification. In this research paper, we introduce a novel methodology that integrates the Active Gradient Deep Convolutional Neural Network (AG-DCNN) architecture with the Red Spider Optimization (RSO) algorithm. Extensive evaluations conducted on well-established retinal image datasets such as DRIVE, STARE, CHASE DB1, HRF, DRISHTI-GS, and RFMiD illustrate substantial enhancements in classification accuracy, sensitivity, and specificity when compared to conventional approaches. The AG-DCNN with RSO exhibits superior performance and remarkable generalization abilities across diverse datasets. This novel approach not only enhances the latest advances in retinal image classification but also holds promise for early disease diagnosis and improved patient care. Through extensive examinations, our proposed method consistently outperforms established techniques, establishing itself as the benchmark in retinal fundus image classification. The integration of AG-DCNN and RSO showcases its efficacy and potential impact in advancing the field of medical image analysis. [ABSTRACT FROM AUTHOR]

With the wide application of deep learning in anomaly detection (AD), industrial vision AD has achieved remarkable success. However, current AD usually focuses on anomaly localization and rarely investigates anomaly classification. Furthermore, anomaly classification is currently requested for quality management and anomaly reason analysis. Therefore, it is essential to classify anomalies while improving the accuracy of AD. This paper designs a novel multi-classification AD (MCAD) framework to achieve high-accuracy AD with an anomaly classification function. In detail, the proposal model based on relational knowledge distillation consists of two components. The first one employs a teacher–student AD model, utilizing a relational knowledge distillation approach to transfer the interrelationships of images. The teacher–student critical layer feature activation values are used in the knowledge transfer process to achieve anomaly detection. The second component realizes anomaly multi-classification using the lightweight convolutional neural network. Our proposal has achieved 98.95, 96.04, and 92.94% AUROC AD results on MNIST, FashionMNIST, and CIFAR10 datasets. Meanwhile, we earn 97.58 and 98.10% AUROC for AD and localization in the MVTecAD dataset. The average classification accuracy of anomaly classification has reached 76.37% in fifteen categories of the MVTec-AD dataset. In particular, the classification accuracy of the leather category has gained 95.24%. The results on the MVTec-AD dataset show that MCAD achieves excellent detection, localization, and classification results. [ABSTRACT FROM AUTHOR]

Automatic handwritten character recognition plays a significant role in various applications across multiple fields. With the growing interest in automatic handwriting recognition and the advancement of deep learning methods, researchers have achieved significant improvements in the development of English handwriting recognition methods. However, the recognition of Arabic handwriting has received insufficient attention. In this paper, a novel "DeepAHR" model is presented to accurately and efficiently recognize Arabic handwritten characters using deep learning techniques. The "DeepAHR" model is based on a convolutional neural network (CNN) and is trained using two recent public datasets: Hijaa and Arabic handwritten characters dataset (AHCD). The overall accuracies of the proposed model were 98.66% and 88.24% on the AHCD and Hijaa datasets, respectively.The experimental results showed that DeepAHR outperformed state-of-the-art methods in the literature. These promising results provide evidence of the successful use of the DeepAHR model for recognizing handwritten Arabic characters [ABSTRACT FROM AUTHOR]

The significant increase in online activities in the wake of recent global events has underlined the importance of biometric person authentication on digital platforms. Although many biometric devices may be used for precise biometric authentication, acquiring the necessary technology, such as 3D sensors or fingerprint scanners, can be prohibitively expensive and logistically challenging. Addressing the demands of online environments, where access to specialized hardware is limited, this paper introduces an innovative approach. In this work, by fusing static and dynamic signature data with facial data captured through regular computer cameras, a dataset of 1750 samples from 25 individuals is constructed. Deep learning models, including convolutional neural networks (CNN), long short-term memory (LSTM), gated recurrent unit (GRU), and temporal convolutional networks (TCN), are employed to craft a robust multi-classification model. This integration of various deep learning algorithms has demonstrated remarkable performance enhancements in biometric authentication. This research also underscores the potential of merging dynamic and static biometric features, derived from readily available sources, to yield a high-performance recognition framework. As online interactions continue to expand, the combination of various biometric modalities holds potential for enhancing the security and usability of virtual environments. [ABSTRACT FROM AUTHOR]

The new Coronavirus disease 2019 (COVID-19) is rapidly affecting the world population with statistics quickly falling out of date. Due to the limited availability of annotated Coronavirus X-ray and CT images, the detection of COVID-19 remains the biggest challenge in diagnosing this disease. This paper provides a promising solution by proposing a COVID-19 detection system based on deep learning. The proposed deep learning modalities are based on convolutional neural network (CNN) and convolutional long short-term memory (ConvLSTM). Two different datasets are adopted for the simulation of the proposed modalities. The first dataset includes a set of CT images, while the second dataset includes a set of X-ray images. Both of these datasets consist of two categories: COVID-19 and normal. In addition, COVID-19 and pneumonia image categories are classified in order to validate the proposed modalities. The proposed deep learning modalities are tested on both X-ray and CT images as well as a combined dataset that includes both types of images. They achieved an accuracy of 100% and an F1 score of 100% in some cases. The simulation results reveal that the proposed deep learning modalities can be considered and adopted for quick COVID-19 screening. [ABSTRACT FROM AUTHOR]

Male infertility negatively affects the lives of infertile couples. Sperm head morphology is an essential factor in evaluating semen in male infertility. The lack of samples with abnormal sperm heads compared to the normal sperm head samples makes the classification of sperm head images an imbalance classification problem. In comparison with other deep networks such as convolutional neural networks, capsule neural networks (CapsNets) provide an optimal platform for designing imbalanced classification models by considering spatial relationships of features. Also, generative adversarial networks (GANs) help to improve the imbalanced classification of images by producing suitable synthetic samples. In this paper, a new architecture based on CapsNets and GANs is proposed and evaluated for imbalanced classification of human sperm head images. The new proposed conditional generative adversarial capsule network outperformed other deep learning networks in the balanced and imbalanced classification of human sperm head images. Based on the comparison between the general methods for increasing the data and the proposed network, the general methods have less robustness to reducing the amount of data than the proposed network. The presented network performed a balanced classification of sperm head images with 97.8% accuracy. Additionally, the proposed network maintained an accuracy of over 80% up to the ratio of minority to majority class of 1:30, indicating that it performed properly in the imbalanced classification of sperm images. [ABSTRACT FROM AUTHOR]

In these recent years, the world has been faced with fires' outbreak, which represented the most serious problem causing huge casualties and considerable destructions. It is therefore essential to early detect fire in video surveillance scenes accurately and reliably in order to overcome the common weaknesses of the available flame detection methods. Nowadays, exploring the recent deep learning (DL)-based methods within the modern surveillance systems has become a great challenge. Thereby, a novel DL-based approach is introduced for fire images detection in this paper. It is based on a convolutional neural network architecture (CNN), designed from scratch and named Fire Classification Network "FireClassNet." Firstly, the input frames are preprocessed to highlight fire regions. Then, they are fed into the proposed "FireClassNet" in order to train the classification model. The presented network includes small number of layers when compared to existing CNNs, resulting in fewer parameters number. Experiments show the effectiveness of the produced model on the constructed dataset in terms of improving the accuracy which reaches 99.73%. It is also demonstrated that the developed model is able to clearly outperform the related methods and the baseline CNN architectures for fire frames classification. [ABSTRACT FROM AUTHOR]

Detection of small objects is one of the main challenges to be improved in deep learning, mainly due to the small number of pixels and scene's context, leading to a loss in performance. In this paper, we present an optimized approach based on deep object detection models that allow the detection of a higher number of elements and improve the score obtained for their class inference. The main advantage of the presented methodology is that it is not necessary to modify the internal structure of the selected convolutional neural network model or re-training for a specific scene. Our proposal is based on detecting initial regions to generate several sub-images using super-resolution (SR) techniques, increasing the number of pixels of the elements, and re-infer over these areas using the same pre-trained model. A reduced set of windows is calculated in the super-resolved image by analyzing a computed graph that describes the distances among the preliminary object detections. This analysis is done by finding maximal cliques on it. This way, the number of windows to be examined is diminished, significantly speeding up the detection process. This framework has been successfully tested on real traffic sequences obtained from the U.S. Department of Transportation. An increase of up to 44.6% is achieved, going from an average detection rate for the EfficientDet D4 model of 14.5% compared to 59.1% using the methodology presented for the first sequence. Qualitative experiments have also been performed over the Cityscapes and VisDrone datasets. [ABSTRACT FROM AUTHOR]

Deep learning has provided many advantages including the ability to extract features from the voice samples and represents the data in a more decisive mode. In speaker identification, many studies have been done to extract more and more meaningful speaker embeddings to improve the system performance. These models perform quite satisfactorily under clean and unambiguous data conditions but with noise and non-speech segments, the functioning of these systems drastically drops down. For real-world scenarios, class noise is more significant as compared to attribute noise for classification problems. The same instances in a sample with different class labels and different classes with similar instances are referred to as class noise. The non-speech regions in different speaker samples carry the same properties and hinder the classification process. We have observed that this non-speaker information (present in noise and non-speech segments) plays a crucial role in the operation of a system. This paper emphasizes the non-speaker embeddings which are very essential for developing an effective model. But in this direction, sufficient research has not yet been done. In this study, we concentrate on this problem and analyses the effect of non-speaker embeddings on the speaker identification process. We have introduced non-speaker classes to solve this issue by learning the non-speaker parameters. Two state-of-art methods, convolutional neural network (CNN), and SincNet with non-speaker embeddings are analyzed to check the performance. The models are trained on non-speaker classes (like silence and noise) along with individual speaker classes resulting in improved system performance with respect to baseline CNN and the SincNet model. We also compare the performance of our approach with the hand-crafted feature (MFCC and FBANK)-based speaker identification process. Here, also, we observed a substantial amount of improvement in classification by our approach. Learning the non-speaker embeddings improve the system performance because the misclassified non-speaker segments are addressed by our approach. For this analysis, speaker data are taken from the LibriSpeech dataset, and non-speaker data are taken from NOISEX-92(Synthetic noise), TESDHE (Natural noise), and Silent CD (Silence) databases. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNN) have emerged as a dominant deep learning technique in various fields, including image processing, computer vision, and intelligent decision-making on embedded devices. Although the underlying structure is simple, the high computation and memory requirements pose significant challenges. To address this issue, low-precision representations of neurons, inputs, model parameters, and activations have become a promising solution. These reduced-precision models offer scalability in performance, storage, and power efficiency while sacrificing some accuracy. By leveraging reconfigurable hardware such as FPGAs-SoC, deep learning systems can take advantage of low-precision inference engines while achieving the desired accuracy and balancing performance, power consumption, and programmability. Despite the high redundancy and excellent classification accuracy provided by CNN, the increasing model size makes it challenging to execute applications on embedded FPGAs. However, recent studies have shown that high levels of accuracy can still be achieved even when weight and activation are scaled down from floating-point (FP) to binary values using approaches such as quantized neural networks (QNN) and binarized neural networks (BNN). In this paper, we review recent works that have utilized binarization and quantization frameworks to explore design space and automate the building of fully customizable inference engines for image processing on FPGAs. [ABSTRACT FROM AUTHOR]

Emotion recognition (ER) from facial images is one of the landmark tasks in affective computing with major developments in the last decade. Initial efforts on ER relied on handcrafted features that were used to characterize facial images and then feed to standard predictive models. Recent methodologies comprise end-to-end trainable deep learning methods that simultaneously learn both, features and predictive model, where the most successful models are based on convolutional neural networks (CNNs). While these models have excelled at this task, they still fail at capturing local patterns that emerge in the learning process. We hypothesize these patterns could be captured by variants based on locally weighted learning. Specifically, in this paper we propose a CNN-based architecture enhanced with multiple branches formed by radial basis function (RBF) units that aims at exploiting local information at the final stage of the learning process (i.e., in the layers close to the output layer). Intuitively, these RBF units capture local patterns shared by similar instances using an intermediate representation, then the outputs of the RBFs are feed to a softmax layer that exploits this information to improve the predictive performance of the model. This feature could be particularly advantageous in ER as cultural / ethnicity differences may be potentially identified by the local units. We evaluate the proposed method in several ER datasets and show the proposed methodology achieves state-of-the-art performance in some of them, even when we adopt a pre-trained VGG-Face model as backbone. We show the proposed method outperforms consistently CNNs that do not have the proposed local learning component. Moreover, we found the proposed model is advantageous when training datasets are reduced and when merging images coming from different distributions (e.g., combining two ER datasets). We show it is the incorporation of local information what makes the proposed model competitive. [ABSTRACT FROM AUTHOR]

Learning is an essential requirement for humans, and its means have evolved. Ten years ago, Massive Open Online Courses (MOOCs) were introduced, attracting many interests and learners. MOOCs provide forums for learners to interact with instructors and to express any problems they encounter in the educational process. However, MOOCs have a high dropout rate due to the difficulties of following up on learners' posts and identifying the urgent ones to react quickly. This research aims to assist instructors in automatically identifying urgent posts, making it easier to respond to such posts rapidly, increasing learner engagement, and improving course completion rate. In this paper, we propose a novel classification model for identifying urgent posts. The proposed model consists of four stages. In the first stage, the post-text is code-encoded and vectorized using a pre-trained BERT model. In the second stage, a novel feature aggregation model is proposed to reveal data-based relationships between token features and their representation in a higher-level feature. In the third stage, a novel model based on convolutional neural networks (CNNs) is proposed to reveal the meaning of a text context more accurately. In the last stage, the extracted composite features are used to classify the text of the post. Several experimental studies were conducted to get the best performance of the proposed stages of the system. The experimental results demonstrated the architectural efficiency of the proposed feature aggregation and multiple CNN models, as well as the accuracy of the proposed system compared to the current research. [ABSTRACT FROM AUTHOR]