Research on the classification of CT images of COVID-19 by pre-training model based on self-attention mechanism

Nowadays, with the outbreak of COVID-19, the prevention and treatment of COVID-19 has gradually become the focus of social disease prevention, and most patients are also more concerned about the symptoms. COVID-19 has symptoms similar to the common cold, and it cannot be diagnosed based on the symptoms shown by the patient, so it is necessary to observe medical images of the lungs to finally determine whether they are COVID-19 positive. As the number of patients with symptoms similar to pneumonia increases, more and more medical images of the lungs need to be generated. At the same time, the number of physicians at this stage is far from meeting the needs of patients, resulting in patients unable to detect and understand their own conditions in time. In this regard, we have performed image augmentation, data cleaning, and designed a deep learning classification network based on the data set of COVID-19 lung medical images. accurate classification judgment. The network can achieve 95.76% classification accuracy for this task through a new fine-tuning method and hyperparameter tuning we designed, which has higher accuracy and less training time than the classic convolutional neural network model. © 2023 SPIE.

Mobile Intellectual Property Marketplace Model for Commercialization of Intellectual Property Rights

This study discusses the development of the intellectual property (IP) marketplace model based on mobile location-aware computing. Referring to statistics released by the Directorate General of Intellectual Property, there has been a growth in the number of intellectual property rights (IPR) applications in recent years, even during the Covid-19 pandemic. On the other hand, after IPR protection, the commercialization of IPR is one of the pillars of the IP system. Nevertheless, research institutions such as LIPI/BRIN indicate that the potential for commercializing IPR is still low. Furthermore, the opportunity is that cellular networks have covered almost all parts of Indonesia, and there has been significant growth in smartphone users. The method utilized in this research is prototyping. This research results from an IP marketplace model based on mobile location-aware computing in Indonesia. Using the smartphone user's location, contextual IPR information from the user's location related to IPR will enter their smartphone. The experimental results indicate that the application can display a list of IPR information according to the smartphone user's location. Furthermore, the search feature can forage IPR listing information based on user queries. © 2022 IEEE.

Tulsi Leaf Disease Detection using CNN

Since the start of the covid 19 pandemic, a wide range of medications have been produced and are currently being utilized to treat the disease. Tulsi, in addition to all of the chemical-based medications, is an herbal therapy that is particularly effective in the treatment of this ailment. Tulsi has been used to heal ailments and infections for millennia, particularly in India. Because we use tulsi for medicinal purposes, it's vital to monitor its health in order to reap the full benefits of its herbal properties. Plant diseases harm the health and growth of the plant. Disease detection in plants is crucial so that it can be treated before it spreads throughout the plant. To detect illnesses in tulsi leaves, we propose employing a model based on convolution neural networks. Image processing and CNN are widely employed. The prepared model extracts the image's key features and categorizes it into different disorders. The model has a 75 percent accuracy rate. © 2022 IEEE.

Evolutionary game analysis of crowding risk control strategies during the COVID-19 Pandemic

To discuss the decision-making scheme of crowding risk management during the COVID-19 pandemic, this paper constructs an evolutionary game model based on the changes of pedestrian and government strategies, and simulates the strategy selection under different states. The results show that under the condition of pedestrian rationality, when the difference between the benefits and costs of the government's active response strategy is less than the benefits of inaction, the government will choose the strategy of inaction. If the benefit of rational action is less than the additional benefit of irrational action, pedestrians will choose irrational action. By establishing the replication dynamic equations of governments and pedestrians, the stability strategy of the system is analyzed. It is found that the values of R1, R2, R3, R4, R5, C1, C2, C3, C4, C5, C6, C7 will affect the strategy choices of the players, and how to measure the benefits and costs under different circumstances becomes the key to the problem. These findings provide a theoretical basis for the risk control decision of human crowding during the COVID-19 epidemic. © 2023 SPIE.

Research on the spread of infectious diseases based on computer simulation

The virus has a big impact on the whole world. The new Coronavirus has a great impact on everyone's life and will even lead to changes in the world pattern. Because of the virus, society is not functioning properly, the recession, people's expectations of economic development are falling. Trains and planes were suspended in some areas. In this paper, computer is used to simulate SIR model, based on system dynamics, to study the spread of infectious diseases. The SIR model passes reality and limit tests. On the basis of the original model, supplementing the original model, isolation and vaccination can effectively stop the spread of the virus. It can slow the outbreak of the virus and reduce the number of infected people. Panic comes from the unknown, and our confidence in defeating the 2019-nCoV virus comes from our scientific base. © 2022 IEEE.

Epidemics: Some Preliminary Results

This chapter presents a novel model based on the operatorial approach for the spread of infections in a healthy population. The model efficiently describes the interactions between healthy and infected populations, and their transformation into recovered or deceased individuals. We apply our model to real situations, showing the efficacy of our method by analyzing Chinese data for SARS-2003 and COVID-19. Our model is in good agreement with the long-term behavior of the diseases, particularly in determining the number of infected and deceased individuals over time. Additionally, we demonstrate how the model can be easily adapted to account for lockdown measures, resulting in a drastic reduction of the number of infected individuals. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Transient zonal modeling of SARS-CoV-2 airborne infection risk for incomplete mixing air distribution methods

A widely used analytical model to quantitatively assess airborne infection risk is the Wells-Riley model based on the assumption of complete air mixing in a single zone. This study aimed to extend the Wells-Riley model so that the infection risk can be calculated in spaces where complete mixing is not present. This is done by evaluating the time-dependent distribution of infectious quanta in each zone and by solving the coupled system of differential equations based on the zonal quanta concentrations. In conclusion, this study shows that using the Wells-Riley model based on the assumption of completely mixing air may overestimate the long-range airborne infection risk compared to some high-efficiency ventilation systems such as displacement ventilation, but also underestimate the infection risk in a room heated with warm air supplied from the ceiling. © 2022 17th International Conference on Indoor Air Quality and Climate, INDOOR AIR 2022. All rights reserved.

An Enhanced Convolution Neural Network Model based on Depthwise-separable Convolutions for Pneumonia Classification

This paper is an attempt to customize a lightweight model to classify pneumonia images by integrating depthwise-separable convolutions with typical CNN model, and focus on the performance of DSCNN in comparison with typical CNN model based on X-ray images. The experimental result shows that in our four-layer structure, DSCNN reduce around 50,000 parameters compared to CNN. But DSCNN had a relative low recall on COVID-19(89.23%). However, with proper means of optimization such as focal loss and data augmentation, there was a slight increase in test accuracy of DSCNN(from 95.25% to 96.14%), and a significant increase in recall on COVID-19(from 89.23% to 94.61%). And this model also performed well on the rest two labels. © 2022 IEEE.

A COVID-19 Detection Model Based on Convolutional Neural Network and Residual Learning

A model that can obtain rapid and accurate detection of coronavirus disease 2019 (COVID-19) plays a significant role in treating and preventing the spread of disease transmission. However, designing such a model that can balance the detection accuracy and weight parameters of memory well to deploy a mobile device is challenging. Taking this point into account, this paper fuses the convolutional neural network and residual learning operations to build a multi-class classification model, which improves COVID-19 pneumonia detection performance and keeps a trade-off between the weight parameters and accuracy. The convolutional neural network can extract the COVID-19 feature information by repeated convolutional operations. The residual learning operations alleviate the gradient problems caused by stacking convolutional layers and enhance the ability of feature extraction. The ability further enables the proposed model to acquire effective feature information at a low cost, which can make our model keep small weight parameters. Extensive validation and comparison with other models of COVID-19 pneumonia detection on the well-known COVIDx dataset show that (1) the sensitivity of COVID-19 pneumonia detection is improved from 88.2% (non-COVID-19) and 77.5% (COVID-19) to 95.3% (non-COVID-19) and 96.5% (COVID-19), respectively. The positive predictive value is also respectively increased from 72.8% (non-COVID-19) and 89.0% (COVID-19) to 88.8% (non-COVID-19) and 95.1% (COVID-19). (2) Compared with the weight parameters of the COVIDNet-small network, the value of the proposed model is 13 M, which is slightly higher than that (11.37 M) of the COVIDNet-small network. But, the corresponding accuracy is improved from 85.2% to 93.0%. The above results illustrate the proposed model can gain an efficient balance between accuracy and weight parameters. © 2023 Tech Science Press. All rights reserved.

Boosting Masked Face Recognition with Multi-Task ArcFace

In this article, we tackle the recognition of faces wearing surgical masks. Surgical masks have become a necessary piece of daily apparel because of the COVID-19-related worldwide health problem. Modern face recognition models are in trouble because they were not made to function with masked faces. Furthermore, in order to stop the infection from spreading, apps capable of detecting if the individuals are wearing masks are also required. To address these issues, we present an end-to-end approach for training face recognition models based on the ArcFace architecture, including various changes to the backbone and loss computation. We also use data augmentation to generate a masked version of the original dataset and mix them on the fly while training. Without incurring any additional computational costs, we modify the chosen network to output also the likelihood of wearing a mask. Thus, the face recognition loss and the mask-usage loss are merged to create a new function known as Multi-Task ArcFace (MTArcFace). The conducted experiments demonstrate that our method outperforms the baseline model results when faces with masks are considered, while achieving similar metrics on the original dataset. In addition, it obtains a 99.78% of mean accuracy in mask-usage classification. © 2022 IEEE.

Simulation Analysis of Infectious Disease Trend Based on Improved SEIR Model

Scholars have carried out a lot of research in the field of using data processing methods to analyze the evolution characteristics and development trends of infectious diseases. The research on data model method is more in-depth, that is, according to the specific characteristics of infectious diseases, suitable data models are designed and combined with different parameters to analyze infectious diseases, mainly including infectious disease data models based on statistical theory or dynamic theory. The former is mostly used in the case of insufficient initial data. Local analysis is carried out by means of a priori or assumptions to achieve global prediction. The latter mainly includes SIR model, complex network model, and cellular automata model. SIR model is the most in-depth research. Scholars have constructed or optimized Si model, SIS model, SEIR model, IR model, and other derivative models based on SIR model in combination with the characteristics of viruses. In this paper, the data source is Wuhan epidemic information released by Health Commission of Hubei Province. Combined with the specific characteristics of COVID-19, the traditional dynamic propagation model is optimized, and an improved SEIR model is constructed. The results of the improved SEIR model are in good agreement with the actual epidemic trend in Wuhan. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Improved SEIR Model Based on Recovery Rate Optimization to Predict COVID-19

When using the traditional SEIR infectious disease model to predict the trend of novel coronavirus pneumonia epidemic, numerous initial parameters need to be tuned, and the parameters cannot change over time during the prediction process, which reduces the accuracy of the model. Firstly, thesis used a logistic model to preprocess the SEIR model parameters and proposed a SEIR model based on time series recovery rate optimization with a new parameter of effective immunity rate. Secondly, the model was trained with epidemic data from domestic and foreign provinces and cities, and the usability of the model was demonstrated experimentally, and the mean absolute percentage error (MAPE) and goodness of fit (R2) were used to compare with other models, which proved the superiority of the model prediction and indicated further research directions. © 2022 IEEE.

Construction and Application of Bid Fraud Prediction Model Based on AdaBoost Algorithm

In recent years, with the development of Internet big data technology and e-commerce platform, many active offline transaction methods have gradually shifted to online. Online auctions have come a long way due to COVID-19, but bidding fraud has seriously disrupted the health of the industry. In this paper, the AdaBoost model is used to build a bidding fraud prediction model, and the prediction performance of the model is verified by data experiments, and it is found that it has a high accuracy for identifying bidding fraud. At present, there are few prediction models for bidding fraud, and it has broad development prospects. © 2022 IEEE.

Fake News Detection System: An implementation of BERT and Boosting Algorithm

On social media, false information can proliferate quickly and cause big issues. To minimize the harm caused by false information, it is essential to comprehend its sensitive nature and content. To achieve this, it is necessary to first identify the characteristics of information. To identify false information on the internet, we suggest an ensemble model based on transformers in this paper. First, various text classification tasks were carried out to understand the content of false and true news on Covid-19. The proposed hybrid ensemble learning model used the results. The results of our analysis were encouraging, demonstrating that the suggested system can identify false information on social media. All the classification tasks were validated and shows outstanding results. The final model showed excellent accuracy (0.99) and F1 score (0.99). The Receiver Operating Characteristics (ROC) curve showed that the true-positive rate of the data in this model was close to one, and the AUC (Area Under The Curve) score was also very high at 0.99. Thus, it was shown that the suggested model was effective at identifying false information online. © 2023, EasyChair. All rights reserved.

Micromodel and Analysis of COVID-19 Based on Hypergraph

In view of the fact that the existing propagation models ignore the influence of different fields and different virus variants on individual infection, and the classical propagation models only describe the macroscopic situation of virus transmission, which cannot be specific to individual cases, this paper proposes 67ya microscopic virus propagation model based on hypergraph (HC-SIRS). Firstly, the concept of hypergraph is used to divide different fields of individuals into corresponding hyperedges. Based on different contact probabilities of each hyperedge, the contact probability matrix is formed to relate the contact between individuals. The individual infection probability of micro-virus propagation model based on hypergraph is deduced, and the corresponding differential equation is established. Secondly, the basic regeneration number and its characteristics of the model are derived. The upper bound of the basic regeneration number of the model is less than or equal to that of the classical SIRS model, indicating that the virus is more difficult to spread in this model. In fact, the different fields people live in and the different personal constitutions have a certain impact on the spread of the virus. The model is more comprehensive, so it is more suitable for simulating the spread of the virus in theory. Finally, the COVID-19 data of Diamond Princess and two cities in China are used for simulation experiments, and the mean absolute error(MAE) is used as the evaluation standard. The results showed that HC-SIRS could well simulate the spread of COVID-19. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

A New Pneumonia Detection Model Based on Transformer with Improved Self-Attention Mechanism

Clinical diagnosis based on computed tomography (CT) could be used, as part of diagnosis standard of COVID-19 pneumonia. Addressing the problem that accuracy of CT-based traditional pneumonia classification diagnosis models is relatively low when employed for classification of community-acquired pneumonia (CP), COVID-19 pneumonia (NCP) and normal cases, a new network model is proposed which combines application of Swin Transformer and multi-head axial self-attention (MASA) mechanism, to analyze CT images and make intelligence-assisted diagnosis. The method in detail is to partially replace traditional multi-head self-attention (MSA) mechanism in encoders of Swin Transformer by MASA. The improved model is applied to train and test on commonly used pneumonia CT dataset CC-CCII. The results show that the proposed network outperforms traditional networks ResNet50 and Vision Transformer in indicators of accuracy, sensitivity and F1-measure. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Rapid Face Mask Detection and Person Identification Model Based on Deep Neural Networks

As COVID-19 has been constantly getting mutated and in three or four months a new variant gets introduced to us and it comes with more deadly problems. The things that prevent us from getting Covid are getting vaccinated and wearing a face mask. In this paper, we have implemented a new Face Mask Detection and Person Recognition model named Insight face which is based on SoftMax loss classification algorithm Arc Face loss and named it as Rapid Face Detection and Peron Identification Model based on Deep Neural Networks (RFMPI-DNN) to detect face mask and person identity rapidly as compared to other models available. To compare our new model, we have used previous MobileNet_V2 model and face recognition module for effective comparison on the basis of time. The proposed model implemented in the system has outperformed the model compared in this paper in every aspect. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Light Weighted Model Based on Convolutional Neural Network for Detection of COVID-19 Using Chest X-Ray Images

Since the discovery of COVID-19, new variants have been emerging. The latest in this series is BA.2, which is the subvariant of omicron and is more transmissible than the previous ones. People infected with this virus must be diagnosed at the earliest to provide the needed clinical attention. Radiological images of the chest are crucial in diagnosing the severity of BA.2 infection RT-PCR (Reverse Transcription-Polymerase Chain Reaction) is one of the approved diagnostics for COVID-19. Moreover, it takes time to give the result as compared to imaging techniques like X-ray and CT scans. Deep learning methods offer a clearer understanding and assist in extracting important data from X-ray images. In the absence of clinical assistance, this research emphasises the benefits of employing deep learning to ascertain the infection's existence. We present a light-weighted convolutional neural network-based deep learning binary classification model in this paper. Dataset consists of 16808 publically available images. The accuracy of our model is 98.76% which is effective to diagnose such patients. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Pre-Trained Xception Model-based COVID Detection using CXR Images

Providing the essential medical resources for COVID-19 diagnosis is a challenge on a worldwide scale. They should be cutting-edge tools that can rapidly identify and analyze the virus using a sequence of tests, and it should also be reasonably priced. A chest X-ray scan is an excellent screening tool, but if several exams are taken, the images produced by the devices must be reviewed swiftly and accurately. Predicting the progression of COVID-19-induced longitudinal lung parenchymal ground glass and the resulting consolidation of pulmonary opacity is highly challenging. Sometimes, COVID-19 will cause pulmonary opacity to consolidate, giving it a rounded appearance and a distribution on the periphery of the lungs. This study introduces the Xception model for predicting COVID-19 in chest x-rays. Chest x-rays may predict the presence of Covid-19 with an accuracy of around 97.83%. © 2022 IEEE

Korean Language NLP Model Based Emotional Analysis of LGBTQ Social Media Communities

Sexual minorities are increasingly gaining social visibility and legal rights guarantees at the constitutional level across much of the world, from South America, the United States, and Europe to Japan, Taiwan, and Thailand. At the same time, the COVID-19 pandemic has brought on significant mental health challenges for the public due to accompanying social and economic impact and measures, most of them adverse. Given pre-existing studies highlighting the minority demographic's vulnerability to depression and other mental health symptoms, and the increasing availability of accessible NLP tools, datasets, and models, this paper uses an emotional classification model to analyze emotional trends in queer communities on social media. Using KoBERT with a pre-labelled dataset containing some forty thousand scraped social media posts labelled with emotions, patterns of emotional expression on Twitter in the queer community is revealed. Resulting data provided a validation of the viability of this method of analyzing trends in negative and positive emotional expression as well as the impact COVID-19 had on online queer communities in early 2020 but revealed limitations. © 2023 IEEE.