Search

In recent years, the research pertaining to speech recognition technology in the Kazakh language has gained significant importance. This is due to the increasing demand for natural language processing applications in the region where Kazakh is predominantly spoken. Thus, there exists an urgent requirement for precise and dependable speech recognition systems. The research study examines the application of sophisticated deep learning methodologies, such as Natural Language Processing (NLP) and Hidden Markov Model (HMM), in facilitating speech recognition for the Kazakh language. Additionally, the investigation delves into how various techniques, including data preprocessing, acoustic modeling, and language modeling, can aid in devising effective speech recognition systems. The article deliberates on the feasible uses of speech recognition technology in the geographic area where Kazakh language is spoken and outlines its future research prospects. The investigation underscores the significance of persistent inquiry in this realm to confront distinctive obstacles encountered in creating speech recognition systems for languages with restricted resources. [ABSTRACT FROM AUTHOR]

The growth of artificial intelligence combined with the collection and storage of large amounts of data in the electronic medical record collection has created an opportunity for orthopedic research and translation into the clinical environment. Machine learning (ML) is a type of artificial intelligence tool well suited for processing the large amount of available data. Specific areas of ML frequently used by orthopedic surgeons performing total joint arthroplasty include tabular data analysis (spreadsheets), medical imaging processing, and natural language processing (extracting concepts from text). Previous studies have discussed models able to identify fractures in radiographs, identify implant type in radiographs, and determine the stage of osteoarthritis based on walking analysis. Despite the growing popularity of ML, there are limitations including its reliance on "good" data, potential for overfitting, long life cycle for creation, and ability to only perform one narrow task. This educational article will further discuss a general overview of ML, discussing these challenges and including examples of successfully published models. [Display omitted] [ABSTRACT FROM AUTHOR]

Accurate shelf-life estimation of food products has the potential to improve the safety, reliability, and sustainability of food supply chains. Here, a scalable system for accurately indicating the freshness-related information of salmon for consumers was developed and characterized. The system comprises two parts, an optically changeable colorimetric sensor by dye incorporated with sol-gel functionalized paper sensor and a deep convolutional neural network (DCNN)-based freshness estimation system based on images. The paper sensor was produced based on paper and immersed, coated, or printed with (3-aminopropyl) triethoxysilane (APTMS) and tetraethyl orthosilicate (TEOS) sol-gel particles and indicators (e.g., pH indicator, ammonia gas indicator, oxidation indicator, and other volatile organic compounds indicators). After being augmented by machine learning, the limit of detection reaches as low as 17.1 ppm for ammonia detection and the accuracy achieved up to 99.2% for salmon freshness estimation, which satisfied the application in food supply chains. The system balanced the trade-off between detection sensitivity, stability, and cost with the ability for scalable batch production. [ABSTRACT FROM AUTHOR]

• We propose a semi-automatic annotation pipeline to obtain high-quality annotations for scientific papers. • We use a transformer-based generative model to generate both high quality layouts of papers and their annotations. • We populate layout regions with synthetically generated content exploiting a layout-agnostic technique. [Display omitted] Document layout analysis is an important task to extract information from scientific literature. Deep-learning solutions for document layout analysis require large collections of training data that are not always available. We generate a large number of synthetic pages to subsequently train a neural network to perform document object detection. The proposed pipeline allows users to deal with less common layouts for which it is not easy to find large annotated datasets. High-quality annotations for a small collection of papers are obtained through a semi-automatic approach. Then, a generative model, based on LayoutTransformer, is used to generate plausible layouts that are subsequently populated with random information to perform data augmentation. We evaluate the proposed method considering scientific articles with two different types of layouts: double and single columns. For double-column papers, we improve detection by 1% starting from 385 manually annotated scientific articles. For single-column papers, we improve detection by 49% starting from 218 articles. [ABSTRACT FROM AUTHOR]

[Display omitted] • Deep learning enabled predictive nucleic acid amplification tests analysis. • On-site fluorescence signal processing powered by cloud computing. • Accurate prediction can be obtained using the early 22.5% data. • Approach can be universally extended to other areas of biomedical research. Pandemics such as COVID-19 have exposed global inequalities in essential health care. Here, we proposed a novel analytics of nucleic acid amplification tests (NAATs) by combining paper microfluidics with deep learning and cloud computing. Real-time amplifications of synthesized SARS-CoV-2 RNA templates were performed in paper devices. Information pertained to on-chip reactions in time-series format were transmitted to cloud server on which deep learning (DL) models were preloaded for data analysis. DL models enable prediction of NAAT results using partly gathered real-time fluorescence data. Using information provided by the G-channel, accurate prediction can be made as early as 9 min, a 78% reduction from the conventional 40 min mark. Reaction dynamics hidden in amplification curves were effectively leveraged. Positive and negative samples can be unbiasedly and automatically distinguished. Practical utility of the approach was validated by cross-platform study using clinical datasets. Predicted clinical accuracy, sensitivity and specificity were 98.6%, 97.6% and 99.1%. Not only the approach reduced the need for the use of bulky apparatus, but also provided intelligent, distributable and robotic insights for NAAT analysis. It set a novel paradigm for analyzing NAATs, and can be combined with the most cutting-edge technologies in fields of biosensor, artificial intelligence and cloud computing to facilitate fundamental and clinical research. [ABSTRACT FROM AUTHOR]

Industrial collaborative robots play an essential role in smart manufacturing because they improve productivity while also ensuring workplace safety. However, the development of prognostic and health management systems to ensure the reliability of these robots has been a major challenge due to the lack of fault data. This paper proposed a digital twin scheme based on the fusion of the robot kinematic and dynamic models' information down to the powertrains (i.e., the joints motor, and gear) along with the control algorithms and uncertainty accommodation based upon deep learning. The presented digital twin concept has the potential to propel simulation-based fault prediction. We also highlight and discuss challenges and opportunities around the development of the hybrid digital twin for condition monitoring of industrial collaborative robots. [ABSTRACT FROM AUTHOR]

• A text-content based approach that can reveal the scholarly impact of papers without human domain-specific knowledge. • Modeling a paper's characteristic by leveraging the spatial relations in the representation space. • The simulations based on regression model could reproduce the process of scholarly impact generation. • Introducing the large-scale Pre-Trained Model to address a classical problem in bibliometrics. Citation and citation-based metrics are traditionally used to quantify the scholarly impact of scientific papers. However, for documents without citation data, i.e., newly published papers, the citation-based metrics are not available. By leveraging deep representation techniques, we propose a text-content based approach that may reveal the scholarly impact of papers without human domain-specific knowledge. Specifically, a large-scale Pre-Trained Model (PTM) with 110 million parameters is utilized to automatically encode the paper into the vector representation. Two indicators, τ (Topicality) and σ (Originality) , are then proposed based on the learned representations. These two indicators leverage the spatial relations of paper representations in the semantic space to capture the impact-related characteristics of a scientific paper. Extensive experiments have been conducted on a COVID-19 open research dataset with 1,056,660 papers. The experimental results demonstrate that the deep representation learning method can better capture the scientific content in the published literature; and the proposed indicators are positively and significantly associated with a paper's potential scholarly impact. In the multivariate regression analysis for the potential impact of a paper, the coefficients of σ and τ are 5.4915 (P < 0.001) and 6.6879 (P < 0.001) for next 6 months prediction, 12.9964 (P < 0.001) and 13.8678 (P < 0.001) for next 12 months prediction. The proposed framework may facilitate the study of how scholarly impact is generated, from a textual representation perspective. [ABSTRACT FROM AUTHOR]

With the latest development in information technology, smart campuses have become a topic that attracts attention in the world of education, especially in higher education. The smart campus has the potential to change the way of managing, accessing facil- ities, and teaching and learning process. The smart campus can collaborate various latest technologies such as edge computing, blockchain, deep learning, virtual reality, and other trend recent technologies. Therefore, this research is a guideline for developing an effective and efficient smart campus system with the latest information and communication technology. This paper explains smart campus technology, features, architecture that have been implemented in real life. Thirty-two papers will be analyzed using the Systematic Literature Review (SLR) method. Hopefully, this guideline can be a valuable reference for smart campus designers, researchers, and practitioners interested in developing innovative and efficient smart campus systems. [ABSTRACT FROM AUTHOR]

• Qualitative and quantitative detection of amine gases could be achieved by CSA. • A visible detection of beef freshness using the amine-responsive CSA was proposed. • ResNet34 had the best performance for beef freshness detection based on CSA. • T-SNE could further visualize and understand the classification process of DL. This study developed an innovative approach that combines a colourimetric sensor array (CSA) composed of twelve pH-response dyes with advanced algorithms, aiming to detect amine gases and assess the freshness of chilled beef. With the assistance of multivariate statistical analysis, the sensor array can effectively distinguish five amine gases and enable rapid quantification of trimethylamine vapour with a limit of detection (LOD) of 8.02 ppb and visually monitor the fresh levels of chilled beef. Moreover, the utilization of deep learning models (ResNet34, VGG16, and GoogleNet) for chilled beef freshness evaluation achieved an overall accuracy of 98.0 %. Furthermore, t -distributed stochastic neighbour embedding (t -SNE) visualized the feature extraction process and provided explanations to understand the classification process of deep learning. The results demonstrated that applying deep learning techniques in the process of pattern recognition of CSA can help in realizing the rapid, robust, and accurate assessment of chilled beef freshness. [ABSTRACT FROM AUTHOR]

Pathogens cause various infectious diseases and high morbidity and mortality which is a global public health threat. The highly sensitive and specific detection is of significant importance for the effective treatment and intervention to minimise the impact. However, conventional detection methods including culture and molecular method gravely depend on expensive equipment and well-trained skilled personnel, limiting in the laboratory. It remains challenging to adapt in resource-limiting areas, e.g., low and middle-income countries (LMICs). To this end, low-cost, rapid, and sensitive detection tools with the capability of field testing e.g., a portable device for identification and quantification of pathogens, has attracted increasing attentions. Recently, paper-based microfluidic analytical devices (μPADs) have shown a promising tool for rapid and on-site diagnosis, providing a cost-effective and sensitive analytical approach for pathogens detection. The fast turn-round data collection may also contribute to better understanding of the risks and insights on mitigation method. In this paper, critical developments of μPADs for in-field detection of pathogens both for clinical diagnostics and environmental surveillance are reviewed. The future development, and challenges of μPADs for rapid and onsite detection of pathogens are discussed, including using the cross-disciplinary development with, emerging techniques such as deep learning and Internet of Things (IoT). [Display omitted] • The state-of-the-art on μPAD is critically reviewed. • μPAD enables onsite detections of biochemical markers from molecules to organisms. • Insights is provided highlighting on multidisciplinary aspects. [ABSTRACT FROM AUTHOR]

With the advancement of science and technology, the number of academic papers published each year has increased almost exponentially. While a large number of research papers highlight the prosperity of science and technology, they also give rise to some problems. As we know, academic papers are the most intuitive embodiment of the research results of scholars, which can reflect the level of researchers. It is also the standard for evaluation and decision-making of them, such as promotion and allocation of funds. Therefore, how to measure the quality of an academic paper is very critical. The most common standard for measuring the quality of academic papers is the number of citation counts of them, as this indicator is widely used in the evaluation of scientific publications. It also serves as the basis for many other indicators (such as the h-index). Therefore, it is very important to be able to accurately predict the citation counts of academic papers. To improve the effective of citation counts prediction, we try to solve the citation counts prediction problem from the perspective of information cascade prediction and take advantage of deep learning techniques. Thus, we propose an end-to-end deep learning framework (DeepCCP), consisting of graph structure representation and recurrent neural network modules. DeepCCP directly uses the citation network formed in the early stage of the paper as the input, and outputs the citation counts of the corresponding paper after a period of time. It only exploits the structure and temporal information of the citation network, and does not require other additional information. According to experiments on two real academic citation datasets, DeepCCP is shown superior to the state-of-the-art methods in terms of the accuracy of citation count prediction. [ABSTRACT FROM AUTHOR]

Municipal papers and polyethylene terephthalate (PET) wastes are primary constituents of municipal solid waste (MSW) which necessitates effective disposal. Co-pyrolysis is an efficient transformation technology for the production of valuable gas and liquid fuels. Therefore, co-pyrolysis characteristics, kinetics, gas and liquid products of cellulose and PET mixtures were investigated in a two-stage reactor by multiple characterizations coupling with deep learning approach. The results indicated that there exhibited positive interaction for cellulose and PET co-pyrolysis in terms of mass difference (δ m) and mass loss rate difference (δ r) between experimental and calculated values. The thermal decomposition process and synergistic effects (δ m and δ r) were accurately predicted by deep learning approaches (R 2 =0.9823–0.9993) using temperature, catalyst, mixing ratio and heating rate variables. Feature importance revealed that the temperature, catalyst, heating rate and mixing ratio were in the descending order to affect the remaining mass from TG curves. And the temperature played a predominant role in synergistic effects. The Flynn-Wall-Ozawa (FWO) method was more suitable for fitting the pyrolysis process to obtain activation energy (E a) with higher R 2 (0.94714–0.99640) compared with Kissinger-Akahira-Sunose (KAS) method. When the mixing ratio of cellulose and PET was 1:1, the E a was significantly reduced to 80.73 KJ/mol, and further reduced to 73.04 KJ/mol by introducing the catalyst, as compared to E a of individual cellulose (178.95 KJ/mol) and PET (198.34 KJ/mol). Co-pyrolysis of cellulose and PET led to a reduction of carbohydrates, furans, alcohols, ketones, and phenols, while the content of acids was remarkably increased which was consistent with the observation in FTIR that the absorption peak of C＝O became sharp and strong. The introduction of ZSM-11 resulted in an increase in aromatics and a decrease in acids content owing to aromatization and deoxygenation. This work is expected to offer beneficial guidance for the efficient disposal of MSW and the application of deep learning methods. [Display omitted] • The interactive pathway for catalytic co-pyrolysis of cellulose and PET was proposed. • The activation energy reached the lowest 73.04 KJ/mol when the mixing ratio was 1:1. • Optimized convolutional neural network predicted synergistic effects with R 2 > 0.99. • Feature importance for TG was temperature > catalyst > heating rate > mixing ratio. [ABSTRACT FROM AUTHOR]

During global outbreaks such as COVID-19, regular nucleic acid amplification tests (NAATs) have posed unprecedented burden on hospital resources. Data of traditional NAATs are manually analyzed post assay. Integration of artificial intelligence (AI) with on-chip assays give rise to novel analytical platforms via data-driven models. Here, we combined paper microfluidics, portable optoelectronic system with deep learning for SARS-CoV-2 detection. The system was quite streamlined with low power dissipation. Pixel by pixel signals reflecting amplification of synthesized SARS-CoV-2 templates (containing ORF1ab, N and E genes) can be real-time processed. Then, the data were synchronously fed to the neural networks for early prediction analysis. Instead of the quantification cycle (C q) based analytics, reaction dynamics hidden at the early stage of amplification curve were utilized by neural networks for predicting subsequent data. Qualitative and quantitative analysis of the 40-cycle NAATs can be achieved at the end of 22nd cycle, reducing time cost by 45%. In particular, the attention mechanism based deep learning model trained by microfluidics-generated data can be seamlessly adapted to multiple clinical datasets including readouts of SARS-CoV-2 detection. Accuracy, sensitivity and specificity of the prediction can reach up to 98.1%, 97.6% and 98.6%, respectively. The approach can be compatible with the most advanced sensing technologies and AI algorithms to inspire ample innovations in fields of fundamental research and clinical settings. [Display omitted] • Paper device with deep learning enable portable and rapid nucleic acid amplification. • Accurate prediction can be obtained using partly gathered real-time data. • Deep learning models used by on-chip tests were well-suited to clinical study. • Approach can inspire ample innovations in fields of analytical science. [ABSTRACT FROM AUTHOR]

Smartphone has long been considered as one excellent platform for disease screening and diagnosis, especially when combined with microfluidic paper-based analytical devices (μPADs) that feature low cost, ease of use, and pump-free operations. In this paper, we report a deep learning-assisted smartphone platform for ultra-accurate testing of paper-based microfluidic colorimetric enzyme-linked immunosorbent assay (c-ELISA). Different from existing smartphone-based μPAD platforms, whose sensing reliability is suffered from uncontrolled ambient lighting conditions, our platform is able to eliminate those random lighting influences for enhanced sensing accuracy. We first constructed a dataset that contains c-ELISA results (n = 2048) of rabbit IgG as the model target on μPADs under eight controlled lighting conditions. Those images are then used to train four different mainstream deep learning algorithms. By training with these images, the deep learning algorithms can well eliminate the influences of lighting conditions. Among them, the GoogLeNet algorithm gives the highest accuracy (>97%) in quantitative rabbit IgG concentration classification/prediction, which also provides 4% higher area under curve (AUC) value than that of the traditional curve fitting results analysis method. In addition, we fully automate the whole sensing process and achieve the "image in, answer out" to maximize the convenience of the smartphone. A simple and user-friendly smartphone application has been developed that controls the whole process. This newly developed platform further enhances the sensing performance of μPADs for use by laypersons in low-resource areas and can be facilely adapted to the real disease protein biomarkers detection by c-ELISA on μPADs. [Display omitted] • This deep learning-assisted smartphone platform is unaffected by ambient lighting. • A fully automated "image in, answer out" operation fashion. • A 2048 custom image dataset is used to test 4 mainstream deep learning algorithms. • GoogLeNet provides >97% accuracy in quantitative rabbit IgG testing. • The area under the curve (AUC) is 4% higher than that of conventional methods. [ABSTRACT FROM AUTHOR]

Academic research is crucial to the development of science and technology and is an important factor that affects national strength. When writing an academic research paper, a rhetorical structure is typically used to present the paper's ideas, but this task is quite difficult for junior researchers. To solve this problem, some studies have adopted text mining to assist with the writing, but the existing methods still require human intervention to generate sentences. Recently, due to the increasing maturity of deep learning technology and the ability to address the problem of automatic text generation, progress has been made in this area. The highly complex deep learning operations can correctly generate sequences and find correlations between sequences. When a user provides a few keywords and key sentences, the proposed algorithm can generate an introduction section for the user. The results show that the generated introduction is more coherent, clearer, and more fluent than existing summarization methods. In addition, the method proposed in this study improves the accuracy compared with traditional text extraction methods. The manuscript produced by this study has been evaluated to show that the study can produce a comprehensive introduction compared with previous studies. • The proposed framework can generate a comprehensive introduction based on users' brief inputs. • This paper demonstrates a three-testing-threshold approach to check the content quality. • The introduction generated through this method has less than 3% similarity to the original text by comparison. [ABSTRACT FROM AUTHOR]

Water holding capacity (WHC) plays an important role when obtaining a high-quality pork meat. This attribute is usually estimated by pressing the meat and measuring the amount of water expelled by the sample and absorbed by a filter paper. In this work, we used the Deep Learning (DL) architecture named U-Net to estimate water holding capacity (WHC) from filter paper images of pork samples obtained using the press method. We evaluated the ability of the U-Net to segment the different regions of the WHC images and, since the images are much larger than the traditional input size of the U-Net, we also evaluated its performance when we change the input size. Results show that U-Net can be used to segment the external and internal areas of the WHC images with great precision, even though the difference in the appearance of these areas is subtle. [ABSTRACT FROM AUTHOR]

Nowadays, there has been a rapidly increasing number of scientific submissions in multiple research domains. A large number of journals have various acceptance rates, impact factors, and rankings in different publishers. It becomes time-consuming for many researchers to select the most suitable journal to submit their work with the highest acceptance rate. A paper submission recommendation system is more critical for the research community and publishers as it gives scientists another support to complete their submission conveniently. This paper investigates the submission recommendation system for two main research topics: computer science and applied mathematics. Unlike the previous works (Wang et al., 2018; Son et al., 2020) that extract TF–IDF and statistical features as well as utilize numerous machine learning algorithms (logistics regression and multiple perceptrons) for building the recommendation engine, we present an efficient paper submission recommendation algorithm by using different bidirectional transformer encoders and the Mixture of Transformer Encoders technique. We compare the performance between our methodology and other approaches by one dataset from Wang et al. (2018) with 14012 papers in computer science and another dataset collected by us with 223,782 articles in 178 Springer applied mathematics journals in terms of top K accuracy (K = 1 , 3 , 5 , 10). The experimental results show that our proposed method extensively outperforms other state-of-the-art techniques with a significant margin in all top K accuracy for both two datasets. We publish all datasets collected and our implementation codes for further references. 1 1 https://github.com/BinhMisfit/PSRMTE. • Bidirectional transformer encoders can improve the performance of the paper submission recommendation system. • The Mixture of Transformer Encoders framework shows the efficiency in the paper submission recommendation problem. • Proposed techniques can surpass other recent techniques on two datasets related. [ABSTRACT FROM AUTHOR]

Both the ABO and Rhesus (Rh) blood groups play crucial roles in blood transfusion medicine. Herein, we report a simple and low-cost paper-based analytical device (PAD) for phenotyping red blood cell (RBC) antigens. Using this Rh typing format, 5 Rh antigens on RBCs can be simultaneously detected and macroscopically visualized within 12 min. The proposed Rh phenotyping relies on the presence or absence of hemagglutination in the sample zones after immobilizing the antibodies targeting each Rh antigen. The PAD was optimized in terms of filter paper type, antibodies, and distance of the visualization zone. In this study, the optimal conditions were Whatman filter paper Grade 4; anti-D, –C, -E, -c, and -e antibodies; RBC suspension of 30%; and a visualization zone of 1 cm above the sample zone. The accuracy of simultaneously phenotyping the five Rh RBC antigens in the blood samples (n = 4692) was 99.19%, comparable with the accuracy of the gold-standard tube method used by blood bank laboratories in several regions of Thailand. Furthermore, decision making based on this method can be assisted by deep learning. After implementing a two-stage objective detection algorithm (YOLO v4-tiny) and classification model (DenseNet-201), the ambiguous images (n = 48) were interpreted with 100% accuracy. The PAD integrated with customized-region convolutional neural networks can reduce the interpretation discrepancies in RBC antigen phenotyping in any laboratory. [Display omitted] • Five Rh antigens typing PADs combined with deep learning algorithm for an assisted human-decision. • Rh antigens typing PADs were tested with real blood samples and compared with the standard tube method. • Dataset of photographs following Rh PADs testing were trained and evaluated by the in-house deep learning platform CiRA CORE. • Misinterpretation caused by weak or ambiguous reactions was solved by the CiRA CORE platform. [ABSTRACT FROM AUTHOR]

A paper-based colourimetric assay for the detection of alanine transaminase has been developed. In the presence of alanine transaminase, 2,4-dinitrophenyl hydrazine changes to pyruvate hydrazone leading to a colour change from pale yellow to dark yellow. Reaction conditions were optimized using absorption spectroscopic studies. Hydrophobic patterns on the Whatman chromatographic paper were created by wax printing, and the reagents were drop cast at the reagent zone. On the paper device, the intensity of the yellow colour increases with ALT concentration in the range of 20–140 U/L in human serum. For the quantification of ALT, coloured images were captured using a digital camera and were processed with Image J software. The machine learning approach was also explored for the ALT analysis by training with colour images of the paper device and testing using a cross-validation procedure. The results obtained with real clinical samples on the paper device showed good accuracy of less than 5% relative error with the clinical lab results. Furthermore, the paper device shows high selectivity to ALT in the presence of various interfering species in blood serum with a sensitivity of 0.261 a.u/(U/L), a detection limit of 4.12 U/L, and precise results with an RSD of less than 7%. For the testing of whole blood, a plasma separation membrane was integrated with the patterned paper. [Display omitted] • Colourimetric detection of ALT on paper strip using 2,4-DNPH. • Negligible interference from other biomolecules. • Tested with serum and ALT spiked serum samples. • Plasma separation membrane incorporated strip used for testing of blood samples. • Quantification of ALT on paper strip using ImageJ and Machine learning approach. [ABSTRACT FROM AUTHOR]

• Methodology to automatically generate geo-located waste container maps. • Use of Computer Vision algorithms to detect waste containers. • Automatic division into locations with and without containers in city maps. • Robust model with consistent performance disregarding the container type. • System evaluated in eleven Spanish cities with an average accuracy of 89%. This paper describes the scientific achievements of a collaboration between a research group and the waste management division of a company. While these results might be the basis for several practical or commercial developments, we here focus on a novel scientific contribution: a methodology to automatically generate geo-located waste container maps. It is based on the use of Computer Vision algorithms to detect waste containers and identify their geographic location and dimensions. Algorithms analyze a video sequence and provide an automatic discrimination between images with and without containers. More precisely, two state-of-the-art object detectors based on deep learning techniques have been selected for testing, according to their performance and to their adaptability to an on-board real-time environment: EfficientDet and YOLOv5. Experimental results indicate that the proposed visual model for waste container detection is able to effectively operate with consistent performance disregarding the container type (organic waste, plastic, glass and paper recycling,...) and the city layout, which has been assessed by evaluating it on eleven different Spanish cities that vary in terms of size, climate, urban layout and containers' appearance. [ABSTRACT FROM AUTHOR]

Developing monitoring system for paper industrial wastewater treatment system is an important route for wastewater reuse and recycling from wastewater, which are regarded as effective way for cleaner production. A novel hybrid deep leaning CLSTMA model, which based on sequential fusion convolutional neural network (CNN), long short term memory (LSTM) and attention mechanism (AM), was developed to monitor the water quality in a full-scale paper industrial wastewater treatment system for energy conservation and emissions reduction. The hybrid CLSTMA model for predicting water quality of paper industrial wastewater treatment system was divided into three steps: spatial information fusion by using CNN module, temporal information fusion by using LSTM module and variable weighted calculation by using AM module. Compare with other models (CNN, LSTM and CLSTM models), RMSE of CLSTMA model for the effluent chemical oxygen demand (COD eff) reduced by 23.3–31.55%, MAE of CLSTMA model reduced by 38.89–74.50%, R of CLSTMA model increased by 8.29–11.86%. For the effluent suspended solids (SS eff), compared with CNN and LSTM models, RMSE of CLSTMA model reduced by 10.26% and 9.92%, MAE of CLSTMA model reduced by 5.37% and 3.44%, R of CLSTMA model increased by 15.13% and 37.21%, respectively. While, R of CLSTMA was consistent with CLSTM model, but RMSE and MAE of CLSTMA model reduced by 16.07% and 7.49% than the CLSTM model. Simulation results demonstrate that the proposed CLSTMA model has a great potential in monitoring paper industrial wastewater treatment system for cleaner production. [Display omitted] • A new deep learning model was proposed for real-time estimation of water quality. • The new model was based on convolutional neural network, long-term short-term memory and attention mechanism. • The new model can guarantee wastewater reuse and reduce the wastewater treatment cost. [ABSTRACT FROM AUTHOR]

Emotional intelligence (EI) involves recognizing, regulating, and understanding emotions in oneself and others, enabling effective relationship management. Emotional Intelligence acts as a differentiator in both personal and professional lives. Nevertheless, many people lack the ability of conveying and interpreting human emotions efficiently due to which they fall short in their social and professional lives. However, this can be improved if people know their Emotional Quotient (EQ) score and work towards improving it by implementing certain skills. Although traditional methods for measuring EQ exist, majority of them are manual and are built on self-efficacy based questionnaire which are often susceptible to faking. In order to overcome these lacuna, this paper proposes a digital approach that can be used for an automatic evaluation of EQ of a person via audio visual signal input. The paper devises four major factors self-awareness, self-regulation, self-motivation and social relationships with their weightages that can be utilised for automatic evaluation through the use of technology. The paper also proposes quantifiable parameters for measuring these factors. The proposed criteria were developed with the emergence of deep learning and its applicability in assessing a person's EQ in mind. [ABSTRACT FROM AUTHOR]

The integration of deep learning techniques from the field of image generation into digital rock analysis has resulted in substantial advancements. However, generating high-quality and highly controllable heterogeneous digital rocks still presents challenges and necessities. This paper reports a digital rock reconstruction method using a multi-condition denoising diffusion probabilistic model (MCDDPM). This method has the diffusion and denoising processes inherent in the diffusion model, ensuring high-quality reconstructed samples. Meanwhile, various physical property parameters are embedded as conditions in the network layers of the model at multiple scales, enhancing the conditional control ability of this method. We compared the porosity, pore size distribution, and two-point correlation coefficient under the same conditions, as well as the evaluation indicators of different methods. The comparative results provide convincing evidence for the accuracy of our proposed digital rock reconstruction method. In addition, studies on the control effect of different conditions on reconstructed samples also demonstrate the effectiveness of our strategy. Therefore, MCDDPM can reconstruct reasonable heterogeneous digital rocks based on combinations of physical property conditions, which provides effective technical support for expanding the library of heterogeneous digital rock samples and even achieving zero-shot learning of digital rocks. • The real numerical rock possesses strong heterogeneity, and the method proposed in this paper can effectively reconstruct heterogeneous digital rock. • Compared to other deep learning network models, the reconstruction effect of the method proposed in this paper is at a high level. • The method we propose supports multiple physical property conditions as input. [ABSTRACT FROM AUTHOR]

• This paper utilizes novel connected vehicle (CV) emulated data to predict real-time crash potential for arterials. • The CV emulated data are flexible to obtain. Different speed-related variables are estimated based on the CV emulated data to depict continuous traffic conditions • The results proved the feasibility of using CV emulated data for real-time crash potential prediction • The proposed methods can be applied to other types of vehicles Real-time crash potential prediction could provide valuable information for Active Traffic Management Systems. Fixed infrastructure-based vehicle detection devices were widely used in the previous studies to obtain different types of data for crash potential prediction. However, it was difficult to obtain data in large range through these devices due to the costs of installation and maintenance. This paper introduced a novel connected vehicle (CV) emulated data for real-time crash potential prediction. Different from the fixed devices' data, CV emulated data have high flexibility and can be obtained continuously with relatively low cost. Crash and CV emulated data were collected from two urban arterials in Orlando, USA. Crash data were archived by the Signal for Analytics system (S4A), while the CV emulated data were obtained through the data collection API with a high frequency. Different data cleaning and preparation techniques were implemented, while various speed-related variables were generated from the CV emulated data. A Long Short-term Memory (LSTM) neural network was trained to predict the crash potential in the next 5−10 min. The results from the model illustrated the feasibility of using a novel CV emulated data to predict real-time crash potential. The average and 50th percentile speed were the two most important variables for the crash potential prediction. In addition, the proposed LSTM outperformed Bayesian logistics regression and XGBoost in terms of sensitivity, Area under Curve (AUC), and false alarm rate. With the rapid development of the connected vehicle systems, the results from this paper can be extended to other types of vehicles and data, which can significantly enhance traffic safety. [ABSTRACT FROM AUTHOR]

An Election is a method of selection of individuals to hold the public office in democracy. Ballot is basically a piece of paper that is used to cast vote during election. In ballot paper voting system each voter uses a ballot paper which is not shared and basically it is a paper printed with the name and symbols of the candidates. The Electronic Voting Machine is basically a memory recorder which records the vote casted by the voters. In this paper, main advantages of E-voting systems for country is highlighted. For constructing E-voting systems, every countries need to do great attention to Verification and Validation requirements. In this research, E-voting scheme with face recognition using deep learning technique is proposed. The process of casting vote is accomplished by blockchain technology and blind signature mechanism. The main objective of the proposed scheme is to explore the positive effects of security and safety in online voting system. • The Election is a method of selection of individuals to hold the public office in democracy. • The Electronic Voting Machine is basically a memory recorder which records the vote casted by the voters. • The aim of the research is to E-voting scheme with face recognition using deep learning technique is proposed. • The main objective of the proposed scheme is to explore the positive effects of security and safety in online voting system. [ABSTRACT FROM AUTHOR]

The impact of air pollution on public health is substantial, and accurate long-term predictions of air quality are crucial for early warning systems to address this issue. Air quality prediction has drawn significant attention, bridging environmental science, statistics, and computer science. This paper presents a comprehensive review of the current research status and advances in air quality prediction methods. Deep learning, a novel machine learning approach, has demonstrated remarkable proficiency in identifying complex, nonlinear patterns in air quality data, yet its application in air quality prediction is still relatively nascent. This paper also conducts a systematic analysis and summarizes how cutting-edge deep learning models are applied in air quality prediction. Initially, the historical evolution of air quality prediction methods and datasets is presented. This is followed by an examination of conventional air quality prediction techniques. A thorough comparative analysis of progress made with both traditional and deep learning-based prediction methods is provided. This review particularly focuses on three aspects: temporal modeling, spatiotemporal modeling, and attention mechanisms. Finally, emerging trends in the field of air quality prediction are identified and discussed. [ABSTRACT FROM AUTHOR]

3D printed ceramic parts often have defects due to their inherent brittleness. These defects are of various scale sizes. Detecting these defects, especially with multiple sizes, is a challenging task in the field of detection. Furthermore, there is a lack of surface defect real-time detection systems suitable for industrial applications in this field. Based on this, this paper makes two contributions: the design and development of a real-time defect detection system, and the proposal of a multiscale defect detection method for 3D printed ceramic surfaces within this system. Firstly, the paper establishes the overall structure of the real-time detection system for surface defects on 3D printed ceramic components, and describes the hardware and software components of the system. On this basis, a dataset of ceramic surface defect images is collected and constructed. Then, experimental analyses point out the shortcomings of the You Only Look Once version-5 (YOLOv5) model for multiscale defect detection. To address the shortcomings, the YOLOv5 model is optimized from three aspects, resulting in the Deep separable convolution + residual network-SKNetwork-Efficient Channel Attention Network-YOLOv5 (DepRes-SK-ECA-YOLOv5) model for multiscale defect detection. This model improves the ability to extract and fuse features of defects at different scales. The experimental results show that the DepRes-SK-ECA-YOLOv5 model can achieve 93.5 %, 91.6 %, 94.3 %, and 0.198 s for Precision, Recall, mAP, and Speed for the test set, respectively. Finally, the paper designs the workflow for the system software. The system hardware and system software are integrated to form a real-time detection system. The performance of the detection system is verified through experiments. [ABSTRACT FROM AUTHOR]

Advances in artificial intelligence (AI) and automated robotics will profoundly influence space operations. By utilising machine learning and deep learning approaches, AI-enabled systems may accomplish tasks as well as improve their own performance. These capabilities are useful in the often-remote settings of outer space and will grow in value as automated space operations become more widespread. As AI extends throughout the space domain, automated algorithms will take on many of the roles that have historically been handled by humans. Artificial intelligence is progressing from theory to implementation in the space environment by exposing new satellites and orbital autonomous vehicles to new data. Even though all initial computational parameters are provided, such systems' outputs can be very unpredictable, putting people, property, and the environment at risk. This paper investigates the application of United Nations space treaties, selected regional AI regulations, and various 'soft-law' instruments and industry initiatives focusing on responsible AI system development to space-based AI systems. Following that, reforms are proposed to clarify the practical relationship between AI systems and the international legal regime that governs space, as well as a 'bottom-up' regulatory approach to better facilitate the future development of regulation governing the use of AI by the global space sector. While this work does not purport to provide a conclusive resolution to these multifaceted matters, its objective is to underscore significant obstacles that arise at the convergence of space law and AI, serving as a preliminary foundation for subsequent discussions on this issue. • Advances in AI and automated robotics will have a profound impact on space operations. Automated algorithms will take on roles traditionally handled by humans as AI becomes more widespread in space. • However, the unpredictable outputs of AI systems can put people, property, and the environment at risk, raising questions about liability. • The paper investigates the application of UN space treaties, regional AI regulations, and industry initiatives to space-based AI systems. • Reforms are proposed to clarify the relationship between AI systems and the international legal regime governing space. • A 'bottom-up' regulatory approach is suggested to facilitate future regulation of AI in the global space sector. [ABSTRACT FROM AUTHOR]

Improving generalization ability in multi-robot formation can reduce repetitive training and calculation. In this paper, we study the multi-robot formation problem with the ability to generalize the target position. Since the generalization ability of neural network is directly proportional to spatial dimension, we adopt the strategy of using different networks to solve different objectives, so that the network learning can focus on the learning of one objective to obtain better performance. In addition, this paper presents a distributed deep reinforcement learning method based on soft actor–critic algorithm for solving multi-robot formation problem. At the same time, the formation evaluation assignment function is designed to adapt to distributed training. Compared with the original algorithm, the improved algorithm can get higher reward cumulative values. The experimental results show that the proposed algorithm can better maintain the desired formation in the moving process, and the rotation design in the reward function makes the multi-robot system have better flexibility in formation. The comparison of control signal curve shows that the proposed algorithm is more stable. At the end of the experiments, the universality of the proposed algorithm in formation maintenance and formation variations is demonstrated. [ABSTRACT FROM AUTHOR]

Cancer is a life-threatening ailment characterized by the uncontrolled proliferation of cells. Breast cancer (BC) represents the most highly infiltrative neoplasms and constitutes the primary cause of mortality in the female population due to cancer-related complications. Consequently, the imperative for early detection and prognosis has emerged as a means to enhance long-term survival rates and mitigate mortality. Emerging artificial intelligence (AI) technologies are being utilized to aid radiologists in the analysis of medical images, resulting in enhanced outcomes for individuals diagnosed with cancer. The purpose of this survey is to examine peer-reviewed computer-aided diagnosis (CAD) systems that have been recently developed and utilize machine learning (ML) and deep learning (DL) techniques for the diagnosis of BC. The survey aims to compare these newly developed systems with previously established methods and provide technical details, as well as the advantages and disadvantages associated with each model. In addition, this paper addresses several unresolved matters, areas of research that require further exploration, and potential avenues for future investigation in the realm of advanced computer-aided design (CAD) models utilized in the interpretation of medical images. Furthermore, the integration of Internet of Things (IoT) in BC research and treatment holds immense significance by facilitating real-time monitoring and personalized healthcare solutions. IoT devices, such as wearable sensors and smart implants, enable continuous data collection, empowering healthcare professionals to track patients' vital signs, response to treatment, and overall health trends, fostering more proactive and tailored approaches to BC management. Moreover, the advent of 5G technology in BC applications promises to revolutionize communication speeds and data transfer, enabling rapid and seamless transmission of large medical datasets. This high-speed connectivity enhances the efficiency of remote diagnostics, telemedicine, and collaborative research efforts, ultimately accelerating the pace of innovation and improving patient outcomes in BC care. The present study aims to examine various classifiers utilized in ML and DL methodologies for the purpose of diagnosing BC. Research findings have demonstrated that DL has superior performance compared to standard ML methods in the context of BC diagnosis, particularly when the dataset is extensive. The existing body of research indicates that there are significant gaps in knowledge that need to be addressed in order to enhance healthcare outcomes in the future. These gaps highlight the pressing need for both practical and scientific research in the field. Finally, IoT and 5G will be how they can be used in order to enhance BC detection, treatment and patient care. • Cancer, characterized by uncontrolled cell proliferation, poses a significant threat to human life. • Breast carcinoma is highly invasive and a leading cause of female cancer-related mortality, emphasizing the need for early detection and prognosis. • Emerging AI technologies are assisting radiologists in analyzing medical images, improving cancer diagnosis outcomes. • This survey focuses on recent computer-aided diagnosis (CAD) systems using machine learning and deep learning for breast carcinoma diagnosis. • The paper identifies unresolved research areas and future investigation prospects in advanced CAD models for medical image interpretation. [ABSTRACT FROM AUTHOR]

Renewable energy sources, driven by the global concern for climate change and the urgent need to reduce greenhouse gas emissions, are gaining increasing importance. Among them, solar energy stands out as a crucial renewable source and accurate solar irradiance forecasting is vital for its effective utilisation. Deep learning (DL) techniques have emerged as state-of-the-art solar irradiance forecasting methods in past years. However, traditional models that are widely used have inferior accuracy in forecasting. To bridge this gap, this paper presents an improved state-of-the-art model that leverages the convolutional property to forecast the time series data. We got a reduction of 8.32% in MAE compared to the base model. The paper's findings highlight the potential of DL-based approaches to significantly improve the accuracy of solar irradiance forecasts. [ABSTRACT FROM AUTHOR]

This paper presents a novel machine learning (ML) based approach to predict machining cycle (part program running) times for complex 5-axis machining. Typical 5-axis machining toolpaths consist of short-segmented discrete linear cutter location lines (CL-lines) with simultaneously varying tool center point (TCP) and orientation vectors (ORI). As the 5-axis machine tool NC (numerical control) system tries to interpolate such part programs smoothly, the TCP motion decelerates and accelerates repeatedly causing the actual feedrate to fluctuate. The actual observed (resultant) feedrate can be approximately 30 % lower than the user commanded one. Furthermore, if the toolpath requires the 5-axis machine tool to travel closer to its singular point, or if the workpiece placement on the table is not optimal, actual feedrate is even lowered further. This paper presents two ML-based approaches to accurately predict 5-axis machining toolpaths. The first strategy uses a bidirectional long short-term memory (Bi-LSTM) network to model the machine tool behavior and generates a direct final cycle time estimation for any given part program. This approach only uses the toolpath geometry to predict the cycle time, and for training it only needs the "final" cycle times of similar part programs making it practical on today's shop floors. The second strategy requires individual CL-line processing times for training. In return, it can provide highly accurate cycle time predictions. Both strategies can capture the interpolation dynamics of 5-axis machine tool NC systems accurately. Simulation studies and experimental validations are conducted on modern 5-axis machine tools with varying workpiece placements and interpolation parameters. Proposed approaches have shown to predict cycle times with 90–95 % accuracy on real-life complex 5-axis machining toolpaths. [ABSTRACT FROM AUTHOR]

In document forensics, identifying ink mismatches is crucial for detecting forgeries and determining document authenticity. However, identifying and separating specific inks from paper can be challenging. Hyperspectral imaging can capture unique spectral patterns exhibited by inks composed of different materials, even when their colors appear identical. Hyperspectral document analysis (HSDI) can be employed to authenticate documents by analyzing the ink. Earlier studies had insufficiently accurate black ink detection, so this paper proposes a novel approach using deep learning and pre-processing methods to improve black ink detection accuracy. The proposed approach uses supervised deep learning to identify ink mismatches in hyperspectral document images. The study evaluated the performance of the proposed model using a hyperspectral image dataset consisting of UWA (University of Western Australia) writing ink in both blue and black colors and different types of artificially identical color inks mixed in various ratios to find ink mismatches. The results showed that the proposed system performed better than other systems reported in the literature, improving the average accuracy by up to 0.18% for blue and 0.36% for black. The proposed method accurately detects ink mismatches and identifies various inks based on their unique spectral response, rendering it highly beneficial for applications in document forensics. [ABSTRACT FROM AUTHOR]

This paper introduces a new visual dataset and framework to facilitate computer-vision-based traffic monitoring in high density, mixed and lane-free traffic (TRAMON). While there are advanced deep learning algorithms that can detect and track vehicles from traffic videos, none of the existing systems provides accurate traffic monitoring in mixed traffic. The mixed traffic flows in developing countries often includes the types of vehicles that are not widely known by the existing visual datasets. The computer vision algorithms also face difficulties in detecting and tracking a high density of vehicles that are not following lanes. This paper proposes a large-scale visual dataset of >282,000 labelled images of traffic vehicles, as well as a comprehensive framework and strategy to train common deep-learning-based computer vision algorithms to detect and track vehicles in high density, heterogeneous and lane-free traffic. A systematic evaluation of results shows that TRAMON, the proposed visual dataset and framework, performs well and better than the common visual dataset at all traffic densities. • We develop TRAMON, a large scale visual database and framework, for the detection and tracking of vehicles in high density, mixed and lane-free traffic • TRAMON outperforms popular 'generic' visual datasets such as Microsoft COCO at all traffic densities. • TRAMON can classify 15 types of urban vehicles in mixed traffic conditions. • TRAMON's highly accurate object detection enables us to not only estimate the traffic volume and density but also to track vehicle trajectories without interruption. [ABSTRACT FROM AUTHOR]

Cancer is a condition in which abnormal cells uncontrollably split and damage the body tissues. Hence, detecting cancer at an early stage is highly essential. Currently, medical images play an indispensable role in detecting various cancers; however, manual interpretation of these images by radiologists is observer-dependent, time-consuming, and tedious. An automatic decision-making process is thus an essential need for cancer detection and diagnosis. This paper presents a comprehensive survey on automated cancer detection in various human body organs, namely, the breast, lung, liver, prostate, brain, skin, and colon, using convolutional neural networks (CNN) and medical imaging techniques. It also includes a brief discussion about deep learning based on state-of-the-art cancer detection methods, their outcomes, and the possible medical imaging data used. Eventually, the description of the dataset used for cancer detection, the limitations of the existing solutions, future trends, and challenges in this domain are discussed. The utmost goal of this paper is to provide a piece of comprehensive and insightful information to researchers who have a keen interest in developing CNN-based models for cancer detection. [ABSTRACT FROM AUTHOR]

Large deep learning models are impressive, but they struggle when real-time data is not available. Few-shot class-incremental learning (FSCIL) poses a significant challenge for deep neural networks to learn new tasks from just a few labeled samples without forgetting the previously learned ones. This setup can easily leads to catastrophic forgetting and overfitting problems, severely affecting model performance. Studying FSCIL helps overcome deep learning model limitations on data volume and acquisition time, while improving practicality and adaptability of machine learning models. This paper provides a comprehensive survey on FSCIL. Unlike previous surveys, we aim to synthesize few-shot learning and incremental learning, focusing on introducing FSCIL from two perspectives, while reviewing over 30 theoretical research studies and more than 20 applied research studies. From the theoretical perspective, we provide a novel categorization approach that divides the field into five subcategories, including traditional machine learning methods, meta learning-based methods, feature and feature space-based methods, replay-based methods, and dynamic network structure-based methods. We also evaluate the performance of recent theoretical research on benchmark datasets of FSCIL. From the application perspective, FSCIL has achieved impressive achievements in various fields of computer vision such as image classification, object detection, and image segmentation, as well as in natural language processing and graph. We summarize the important applications. Finally, we point out potential future research directions, including applications, problem setups, and theory development. Overall, this paper offers a comprehensive analysis of the latest advances in FSCIL from a methodological, performance, and application perspective. • In-depth survey on FSCIL methods, applications, performance. • Categorizing FSCIL into five subcategories for clear analysis. • Evaluating FSCIL research on benchmarks for strengths, weaknesses. • FSCIL applications in computer vision, NLP, and graph. [ABSTRACT FROM AUTHOR]

Deep neural networks have achieved outstanding performance in computer vision tasks. Convolutional Neural Networks (CNNs) typically operate in the spatial domain with raw images, but in practice, images are usually stored and transmitted in their compressed representation where JPEG is one of the most widely used encoder. Also, these networks are computationally intensive and slow. This paper proposes performing the learning and inference processes in the compressed domain in order to reduce the computational complexity and improve the speed of popular CNNs. For this purpose, a novel graph-based frequency channel selection method is proposed to identify and select the most important frequency channels. The computational complexity is reduced by retaining the important frequency components and discarding the insignificant ones as well as eliminating the unnecessary layers of the network. Experimental results show that the modified ResNet-50 operating in the compressed domain is up to 70% faster than the spatial-based traditional ResNet-50 while resulting in similar classification accuracy. Moreover, this paper proposes a preprocessing step with partial encoding to improve the resilience to distortions caused by low-quality encoded images. Finally, we show that training a network with highly compressed data can achieve a good classification accuracy with up to 93% reduction in the storage requirements of the training data. [ABSTRACT FROM AUTHOR]

Strain-based structural health monitoring technology has been widely used in the field of transportation. The existing strain damage identification methods have defects such as complex process, lag in state evaluation, and low intelligence. This paper adopts the deep learning method to establish a network model that uses the strain field information to map directly to the damage information, and takes a subway bolster as the engineering background to realise the end-to-end automatic damage identification. Firstly, the problem of damage identification in strain field is described, combined with the idea of fully convolutional network. The basic structure of damage identification network is modularised, and the overall design framework is proposed. Then, the damage simulation method is determined, and the feasibility of using this method to construct a strain field damage dataset is verified. The batch random damage model generation and the random noise signal addition program are coded, and the datasets of the bolster under static/dynamic force are obtained. Finally, the deep learning model is applied to the bolster damage dataset, and a residual module BolRes_Att that integrates spatial attention and channel attention mechanism is proposed. It has better damage identification performance without an increase in model parameters. The average number of faulty elements on the two bolster test sets of the improved damage identification network is 3.02 and 2.92, respectively, accounting for about 0.016% of all elements. The average processing time for a set of data is only 0.014 s. The results show that the deep learning model constructed in this paper can accurately and quickly identify the damage information of elements according to the strain field information. [ABSTRACT FROM AUTHOR]

Highly discriminative feature expression for non-rigid shape recognition is an important and challenging task, which requires both abstract and robust shape descriptors. However, the majority of existing low-level descriptors are designed via hand-crafted, which are sensitive to local changes and larger deformation. To address this issue, this paper proposes a bag of shape descriptor based on unsupervised deep learning and Bag of Words (BoW) for shape recognition. Different from existing pipelines, our method is specially designed to learn high-level and hierarchical shape features from multi-scale context structures. It effectively overcomes obstacles, such as irregular topology, orientation ambiguity, and rigid or non-rigid transformation in the hierarchical learning of contour fragments. Specifically, by adopting an improved decomposing strategy, the shape can be decomposed to a series of valuable contour fragments, results in local to global feature learning. An unsupervised learning framework is also applied to the contour fragment for its feature expression based on the context structure and SSAE (Stack Sparse Auto Encode). In the process of shape representation, a high-level shape dictionary is learned by K-clustering to achieve discriminative feature coding. In addition, to achieve a compact and simplified shape representation, SPM (Spatial Pyramid Matching) is adopted by max-pooling, which effectively incorporates spatial layout information of the given shape. The experiments demonstrate that the proposed method achieves state-of-the-art performance on several public shape datasets comparing with the latest approaches. Our method also obtains high performance under the noisy and occlusion condition. • Our method is specially designed to learn high-level and hierarchical shape features from multi-scale context structures. • An improved decomposing strategy is redesigned to generate valuable contour fragments, results in local to global feature learning. • An unsupervised learning framework is also applied to the contour fragment for its feature expression based on the context structure and SSAE (Stack Sparse Auto Encode). [ABSTRACT FROM AUTHOR]

• Proposing physics-informed DCNN for HFI-SCB temperature field reconstruction. • Proposing Deep MC-QR method to quantify data uncertainty caused by noise. • Proposing HFI-SCB reliability analysis method based on interval multilevel BN. Temperature field reconstruction is essential for analyzing the reliability of a high-density functionally integrated satellite circuit board (HFI-SCB). As a representative deep learning model, the deep convolutional neural network (DCNN) is a powerful tool for reconstructing the HFI-SCB temperature field. However, DCNN needs a lot of labeled data to learn its parameters, which is contrary to the fact that actual satellite engineering can only acquire noisy unlabeled data. Thus, this paper proposes an unsupervised method, i.e., the physics-informed deep Monte Carlo quantile regression method, for reconstructing the HFI-SCB temperature field and quantifying the data uncertainty caused by sensor noise. The proposed method combines a DCNN with known physics knowledge to reconstruct an accurate HFI-SCB temperature field using only monitoring point temperatures. Besides, the proposed method can quantify the data uncertainty by the Monte Carlo quantile regression. Based on the reconstructed temperature field and the quantified data uncertainty, this paper builds an interval multilevel Bayesian Network to analyze the HFI-SCB reliability. Two case studies are used to validate the proposed method. [ABSTRACT FROM AUTHOR]

The paper presents a novel solution to the issue of incomplete regions in 3D meshes obtained through digitization. Traditional methods for estimating the surface of missing geometry and topology often yield unrealistic outcomes for intricate surfaces. To overcome this limitation, the paper proposes a neural network-based approach that generates points in areas where geometric information is lacking. The method employs 2D inpainting techniques on color images obtained from the original mesh parameterization and curvature values. The network used in this approach can reconstruct the curvature image, which then serves as a reference for generating a polygonal surface that closely resembles the predicted one. The paper's experiments show that the proposed method effectively fills complex holes in 3D surfaces with a high degree of naturalness and detail. This paper improves the previous work in terms of a more in-depth explanation of the different stages of the approach as well as an extended results section with exhaustive experiments. [Display omitted] • We perform 3D surface reconstructions using generative inpainting techniques. • 2D representation of a 3D surface geometry based on its curvature. • Application of a general purpose neural network for inpainting. • Our approach does not require dataset nor training time. • Results outperform state-of-the-art quality and naturalness of the reconstructions. [ABSTRACT FROM AUTHOR]

Computed tomography (CT) and magnetic resonance imaging (MRI) are two widely used clinical imaging modalities for non-invasive diagnosis. However, both of these modalities come with certain problems. CT uses harmful ionising radiation, and MRI suffers from slow acquisition speed. Both problems can be tackled by undersampling, such as sparse sampling. However, such undersampled data leads to lower resolution and introduces artefacts. Several techniques, including deep learning based methods, have been proposed to reconstruct such data. However, the undersampled reconstruction problem for these two modalities was always considered as two different problems and tackled separately by different research works. This paper proposes a unified solution for both sparse CT and undersampled radial MRI reconstruction, achieved by applying Fourier transform-based pre-processing on the radial MRI and then finally reconstructing both modalities using sinogram upsampling combined with filtered back-projection. The Primal–Dual network is a deep learning based method for reconstructing sparsely-sampled CT data. This paper introduces Primal–Dual UNet, which improves the Primal–Dual network in terms of accuracy and reconstruction speed. The proposed method resulted in an average SSIM of 0.932±0.021 while performing sparse CT reconstruction for fan-beam geometry with a sparsity level of 16, achieving a statistically significant improvement over the previous model, which resulted in 0.919±0.016. Furthermore, the proposed model resulted in 0.903±0.019 and 0.957±0.023 average SSIM while reconstructing undersampled brain and abdominal MRI data with an acceleration factor of 16, respectively - statistically significant improvements over the original model, which resulted in 0.867±0.025 and 0.949±0.025. Finally, this paper shows that the proposed network not only improves the overall image quality, but also improves the image quality for the regions-of-interest: liver, kidneys, and spleen; as well as generalises better than the baselines in presence the of a needle. [ABSTRACT FROM AUTHOR]

This paper proposes a three-stage online deep learning model for time series based on the ensemble deep random vector functional link (edRVFL). The edRVFL stacks multiple randomized layers to enhance the single-layer RVFL's representation ability. Each hidden layer's representation is utilized for training an output layer, and the ensemble of all output layers forms the edRVFL's output. However, the original edRVFL is not designed for online learning, and the randomized nature of the features is harmful to extracting meaningful temporal features. In order to address the limitations and extend the edRVFL to an online learning mode, this paper proposes a dynamic edRVFL consisting of three online components, the online decomposition, the online training, and the online dynamic ensemble. First, an online decomposition is utilized as a feature engineering block for the edRVFL. Then, an online learning algorithm is designed to learn the edRVFL. Finally, an online dynamic ensemble method, which can measure the change in the distribution, is proposed for aggregating all layers' outputs. This paper evaluates and compares the proposed model with state-of-the-art methods on sixteen time series. [ABSTRACT FROM AUTHOR]

Artificial Intelligence and Machine learning have been widely used in various fields of mathematical computing, physical modeling, computational science, communication science, and stochastic analysis. Approaches based on Deep Artificial Neural Networks (DANN) are very popular in our days. Depending on the learning task, the exact form of DANNs is determined via their multi-layer architecture, activation functions and the so-called loss function. However, for a majority of deep learning approaches based on DANNs, the kernel structure of neural signal processing remains the same, where the node response is encoded as a linear superposition of neural activity, while the non-linearity is triggered by the activation functions. In the current paper, we suggest to analyze the neural signal processing in DANNs from the point of view of homogeneous chaos theory as known from polynomial chaos expansion (PCE). From the PCE perspective, the (linear) response on each node of a DANN could be seen as a 1st degree multi-variate polynomial of single neurons from the previous layer, i.e. linear weighted sum of monomials. From this point of view, the conventional DANN structure relies implicitly (but erroneously) on a Gaussian distribution of neural signals. Additionally, this view revels that by design DANNs do not necessarily fulfill any orthogonality or orthonormality condition for a majority of data-driven applications. Therefore, the prevailing handling of neural signals in DANNs could lead to redundant representation as any neural signal could contain some partial information from other neural signals. To tackle that challenge, we suggest to employ the data-driven generalization of PCE theory known as arbitrary polynomial chaos (aPC) to construct a corresponding multi-variate orthonormal representations on each node of a DANN. Doing so, we generalize the conventional structure of DANNs to Deep arbitrary polynomial chaos neural networks (DaPC NN). They decompose the neural signals that travel through the multi-layer structure by an adaptive construction of data-driven multi-variate orthonormal bases for each layer. Moreover, the introduced DaPC NN provides an opportunity to go beyond the linear weighted superposition of single neurons on each node. Inheriting fundamentals of PCE theory, the DaPC NN offers an additional possibility to account for high-order neural effects reflecting simultaneous interaction in multi-layer networks. Introducing the high-order weighted superposition on each node of the network mitigates the necessity to introduce non-linearity via activation functions and, hence, reduces the room for potential subjectivity in the modeling procedure. Although the current DaPC NN framework has no theoretical restrictions on the use of activation functions. The current paper also summarizes relevant properties of DaPC NNs inherited from aPC as analytical expressions for statistical quantities and sensitivity indexes on each node. We also offer an analytical form of partial derivatives that could be used in various training algorithms. Technically, DaPC NNs require similar training procedures as conventional DANNs, and all trained weights determine automatically the corresponding multi-variate data-driven orthonormal bases for all layers of DaPC NN. The paper makes use of three test cases to illustrate the performance of DaPC NN, comparing it with the performance of the conventional DANN and also with plain aPC expansion. Evidence of convergence over the training data size against validation data sets demonstrates that the DaPC NN outperforms the conventional DANN systematically. Overall, the suggested re-formulation of the kernel network structure in terms of homogeneous chaos theory is not limited to any particular architecture or any particular definition of the loss function. The DaPC NN Matlab Toolbox is available online and users are invited to adopt it for own needs. • Kernel structure reformulation of deep artificial neural networks in terms of homogeneous chaos theory. • Response on each node of the deep network represented through polynomial chaos expansion. • Orthonormal decomposition mitigates the non-optimal representation while processing the neural signal. • Accounting for high-order effects reflecting simultaneous neural impacts. • Analytical estimation of mean and variance of neural signal on each node is provided. [ABSTRACT FROM AUTHOR]

Attention mechanisms have been widely adopted for image captioning because of their powerful performance. In this paper, we propose a Mask Captioning Network (MaC) consisting of an object layer and a background layer to capture the objects and scenes of an image to generate a sentence. To this end, we leverage the Mask RCNN to detect salient regions at the pixel level instead of a bounding box in the object layer. Meanwhile, in the background layer, a CNN model is used to encode the scene features. In addition, MaC is implemented in both LSTM-based and Transformer-based image captioning architectures. We introduce a mask-guided transformer encoder with additional features to enhance the model. Experimental results show that our model significantly outperforms (with a much richer sentence) baseline models and achieves comparable results with state-of-the-art methods on MSCOCO and Flickr30k datasets. • This paper proposes a new image captioning model consisting of the mask and scene layers. • The idea of the mask layer is to eliminate background information to focus on the image objects only. • The scene layer will focus on generating scene features to ensure that the overall meaning of the images is adapted. • Also, our method can be easily integrated into a transformer-based image captioning model to achieve better performance. • Our method obtains comparable/better results across four popular metrics against SOTA in MSCOCO and Flick30K datasets. [ABSTRACT FROM AUTHOR]

Spatial locations of personnel, equipment, and materials are constantly changing as construction projects progress. The dynamic nature of the construction industry affects workers' performance of identifying hazards. Even though a great deal of effort has been made to improve construction safety, the construction industry still witnesses a high accident rate. In order to complement the existing body of knowledge relating to construction safety, this paper integrates Digital Twin (DT), Deep Learning (DL), and Mixed Reality (MR) technologies into a newly developed real-time visual warning system, which enables construction workers to proactively determine their safety status and avoid accidents. Next, system tests were conducted under three quasi-on-site scenarios, and the feasibility was proven in terms of synchronising construction activities over a large area and visually representing hazard information to its users. These evidenced merits of the development testing scenarios can improve workers' risk assessment accuracy, reinforce workers' safety behaviour, and provide a new perspective for construction safety managers to analyse construction safety status. • The proposed system generates simulated virtual construction sites based on the digital twin concept. • The virtual construction sites integrate BIM models and live video data for hazard identification. • Wearable mixed reality devices can present real-time hazard information to individual workers intuitively. • Three preliminary experiments have demonstrated the feasibility of the proposed system. [ABSTRACT FROM AUTHOR]

Sentiment analysis becomes an essential part of every social network, as it enables decision-makers to know more about users' opinions. Arabic sentiment analysis has a small number of datasets relative to English. This paper introduces active learning to the Arabic NLP community and demonstrates its power. Active learning is the process of selecting the data before labelling it to get the best benefits from the labelled data and improve performance. As it takes much time and effort to label large datasets, a better way is to label selected data and get higher performance. This paper applies active learning to the ArSarcasm-v2 dataset, which is labelled for sentiment, sarcasm, and dialect, by using the SAIDS "Sentiment Analysis Informed of Dialect and Sarcasm" model that predicts sarcasm and dialect and then uses them to predict the sentiment of the text. The paper runs multiple active learning experiments with different setups. The paper achieves state-of-the-art performance on ArSarcasm-v2 for sentiment analysis by achieving a 76.71 F1-score using 95% of the training dataset which means that adding some data points makes the performance worse. The paper also demonstrates that by using active learning, we achieve 97%, 99%, and 100% of the highest F1-score reached by all training data using only 10%, 20%, and 27% of the training data respectively. This means that using about one-quarter of the training data attains the same F1-score as using the full training data. As per our knowledge, this is the first paper to introduce active learning to the Arabic NLP community. [ABSTRACT FROM AUTHOR]