Transfer Learning Model in Deep Neural Network for COVID-19 Prediction from CT Images

Deep learning models are often used to process radi-ological images automatically and can accurately train networks' weights on appropriate datasets. One of the significant benefits of the network is that it is possible to use the weight of a pre-trained network for other applications by fine-tuning the current weight. The primary purpose of this work is to employ a pre-trained deep neural framework known as transfer learning to detect and diagnose COVID-19 in CT images automatically. This paper uses a popular deep neural model, ResNet152, as a neural transfer approach. The presented framework uses the weight obtained from the ImageNet dataset, fine-tuned by the dataset used in the work. The effectiveness of the suggested COVID-19 prediction system is evaluated experimentally and compared with DenseNet, another transfer learning model. The recommended ResNet152 transfer learning model exhibits improved performance and has a 99% accuracy when analogized with the DenseNet201 transfer learning model. © 2022 IEEE.

Fine-Tuning ResNet-152V2 for COVID-19 Recognition in Chest Computed Tomography Images

The epidemic Covid-19 has extended to majority of nations. This pandemic is due to a contagious condition 'SARS-CoV-2', was identified by the the International Health association. In order to diagnosis this virus from 2D chest computed tomography (CT) images, we applied three different transfer learning algorithms: $VGG-19, ResNet-152V2$ and a Fine-Tuned version of $ResNet-152V2$. The different transfer learning models are used on three hundred and four exams where 74 are normal cases, 60 are community-acquired pneumonia (CAP) cases and 169 were confirmed corona-virus cases. The best accuracy value is reached by the fine-tuned $ResNet-152v2$ by 75% against 70% for the basic $ResNet-152v2$ and 66% for the $VGG-19$. © 2022 IEEE.

COVID QA Network: A Specific Case of Biomedical Question Answering

COVID-19 crisis has led to an outburst of information that needs to be organized, validated, and made available to the seekers. Despite the rapid growth and success of BERT models in the last 3 years, COVID QA is a difficult task due to the lack of applicable datasets and a relevant language representation. Therefore, this study proposes a transformer-based Question Answering (QA) model for COVID-19 questions from the biomedical domain. Further, explored several datasets, and models required for question type prediction, no-Answer prediction, and answer extraction and transfer learning strategies. It has been demonstrated that the exact match score can be significantly improved with limited amounts of training data from the biomedical domain. Finally, the findings of the study have been summarized as Factoid QA Finetuning Framework (FQFF), which can provide initial direction for domain-specific QA tasks with a limited amount of data. © 2023 IEEE.

Conspiracy Narratives in the Protest Movement Against COVID-19 Restrictions in Germany. A Long-term Content Analysis of Telegram Chat Groups

From the start of the COVID-19 pandemic in Germany, different groups have been protesting measures implemented by different government bodies in Germany to control the pandemic. It was widely claimed that many of the offline and online protests were driven by conspiracy narratives disseminated through groups and channels on the messenger app Telegram. We investigate this claim by measuring the frequency of conspiracy narratives in messages from open Telegram chat groups of the Querdenken movement, set up to organize protests against COVID-19 restrictions in Germany. We furthermore explore the content of these messages using topic modelling. To this end, we collected 822k text messages sent between April 2020 and May 2022 in 34 chat groups. By fine-tuning a Distilbert model, using self-annotated data, we find that 8.24% of the sent messages contain signs of conspiracy narratives. This number is not static, however, as the share of conspiracy messages grew while the overall number of messages shows a downward trend since its peak at the end of 2020. We further find a mix of known conspiracy narratives make up the topics in our topic model. Our findings suggest that the Querdenken movement is getting smaller over time, but its remaining members focus even more on conspiracy narratives. © 2022 Association for Computational Linguistics.

Exploring Language-Interfaced Fine-Tuning for COVID-19 Patient Survival Classification

We present Language-Interfaced Fine-Tuning (LIFT) in application to COVID-19 patient survival classification. LIFT describes translating tabular Electronic Health Records (EHRs) into text inputs for transformer neural networks. We study LIFT with a dataset of 5,371 COVID-19 patients. We focus on the predictive task of survival classification utilizing demographic and medical history features. We begin by presenting information about our dataset. We preface our investigation in text-based transformers by reporting the performances of conventional machine learning models such as Logistic Regression and Random Forest classifiers. We also present the results of a few configurations of tabular input-based Deep Multilayer Perceptron (MLP) networks. 86% of the patients in our database survived in the measured time window. Thus, predictive models are heavily biased to predict that a patient will survive. We emphasize that this problem of Class Imbalance was a major challenge in developing these models. Our balanced sampling strategy from examples in the majority and minority classes is crucial to achieving even reasonable predictive performance. For this reason, we also report performance based on Precision, Recall, and F-score metrics, in addition to Accuracy. Having established baselines with tabular inputs, we then shift our focus to the prompts for translating from tabular to text inputs. We report the performance of 5 prompts. The LIFT model achieves an F-score on the held-out test set of 0.21, slightly behind the Deep MLP with Tabular Features score of 0.23. Both models outperform the Random Forest with Tabular Features at 0.15. We believe that LIFT is a very exciting direction for machine learning in healthcare applications because text-based inputs enables us to take advantage of recent advances in Transfer Learning and Retrieval-Augmented Learning. This study illustrates the effectiveness of converting tabular EHRs to text inputs and utilizing transformer neural networks for prediction. © 2022 IEEE.

Tourism Demand Prediction after COVID-19 with Deep Learning Hybrid CNN–LSTM—Case Study of Vietnam and Provinces

The tourism industry experienced a positive increase after COVID-19 and is the largest segment in the foreign exchange contribution in developing countries, especially in Vietnam, where China has begun reopening its borders and lifted the pandemic limitation on foreign travel. This research proposes a hybrid algorithm, combined convolution neural network (CNN) and long short-term memory (LSTM), to accurately predict the tourism demand in Vietnam and some provinces. The number of new COVID-19 cases worldwide and in Vietnam is considered a promising feature in predicting algorithms, which is novel in this research. The Pearson matrix, which evaluates the correlation between selected features and target variables, is computed to select the most appropriate input parameters. The architecture of the hybrid CNN–LSTM is optimized by utilizing hyperparameter fine-tuning, which improves the prediction accuracy and efficiency of the proposed algorithm. Moreover, the proposed CNN–LSTM outperformed other traditional approaches, including the backpropagation neural network (BPNN), CNN, recurrent neural network (RNN), gated recurrent unit (GRU), and LSTM algorithms, by deploying the K-fold cross-validation methodology. The developed algorithm could be utilized as the baseline strategy for resource planning, which could efficiently maximize and deeply utilize the available resource in Vietnam.

Real Time Mask Detection using Convolutional Neural Network

With the advent of Convolutional Neural Networks, the field of image classification has seen tremendous growth, with various previously impossible applications now being pursued. One such application is face mask detection, which is an important problem to solve, considering recent pandemic. The novelty of this work is the training of YOLO (You Only Look Once) framework for custom object detection, which in this case is face mask, based on some empirical rules for fine-tuning the performance. Also, image classification is proposed to be combined with tracker, in order to implement real world access grant system based on compliance shown by mask wearer. © 2022 IEEE.

Transfer learning-based CNN diagnostic framework for diagnosis of COVID-19 from lung CT images

COVID-19 is an infectious disease that affects the respiratory system. To assist the physician in diagnosing lung disorders from chest CT images various systems have been developed and used. Detection of COVID-19 remains a challenging area of research. The objective of the work is to develop an inductive parameter-transfer learning-based approach for the prediction of COVID-19, pneumonia, from lung CT images. Our proposed approach is built on layer wise and convolution block-wise fine-tuning which designs the CNN architecture highly specific to lung CT image. We implemented the DenseNet201, InceptionV3, Xception, VGG19, and ResNet50 as baseline models. The network architectures are developed to learn feature representation of lung CT images. For the experimental analysis, five datasets are used. From the experimental results, it is inferred that the DenseNet201 model yields higher accuracy of 0.94 for Adam optimizer and 0.93 for the RMSprop optimizer compared to other models.

Prototyping a VR Sandbox for Scene Customization without 3D Authoring Skills

Virtual Reality (VR) and simulation continue po-sitioning as suitable tools for fine-tuning processes otherwise impossible in real life. Such is the case of Aether, a mobile service robot for elderly care developed during the COVID-19 pandemic. Aether's development was negatively impacted due to restrictions placed on accessing long-term care facilities that impeded testing object tracking, elderly tracking, fall detection, and human-robot interactions. Our efforts to maximize Aether's development led us to create a digital twin where the core functionality is replicated to train the machine learning modules to optimize the robot's responses before real-world deployment. However, the digital twin creation requires significant authoring to ensure the virtual environment matches the real one by employing 3D technical artistry skills, which demands a professional knowledgeable in this domain. This paper presents a sandbox prototype for scene customization that allows importing, positioning, scaling, and saving changes for mobile robot simulation. Our preliminary testing of the sandbox has focused on usability to understand how the setting up of the environment is perceived. Preliminary results indicate that the sandbox is usable with improvements pertaining to improving the manipulation of the objects. © 2022 IEEE.

Identifying Similar Questions in the Medical Domain Using a Fine-tuned Siamese-BERT Model

A large number of people search about their health related problems on the web. However, the number of sites with qualified and verified people answering their queries is quite low in comparison to the number of questions being put up. The rate of queries being searched on such sites has further increased due to the COVID-19 pandemic. The main reason people find it difficult to find solutions to their queries is due to ineffective identification of semantically similar questions in the medical domain. For most cases, answers to the queries people ask would be present, the only caveat being the question may be present in a different form than the one asked by the particular user. In this research, we propose a Siamese-based BERT model to detect similar questions using a fine-tuning approach. The network is fine-tuned with medical question-answer pairs and then with question-question pairs to get a better question similarity prediction. © 2022 IEEE.

Transfer learning-based CNN diagnostic framework for diagnosis of COVID-19 from lung CT images

COVID-19 is an infectious disease that affects the respiratory system. To assist the physician in diagnosing lung disorders from chest CT images various systems have been developed and used. Detection of COVID-19 remains a challenging area of research. The objective of the work is to develop an inductive parameter-transfer learning-based approach for the prediction of COVID-19, pneumonia, from lung CT images. Our proposed approach is built on layer wise and convolution block-wise fine-tuning which designs the CNN architecture highly specific to lung CT image. We implemented the DenseNet201, InceptionV3, Xception, VGG19, and ResNet50 as baseline models. The network architectures are developed to learn feature representation of lung CT images. For the experimental analysis, five datasets are used. From the experimental results, it is inferred that the DenseNet201 model yields higher accuracy of 0.94 for Adam optimizer and 0.93 for the RMSprop optimizer compared to other models. © 2023 The Royal Photographic Society.

A MobileNet-based CNN model with a novel fine-tuning mechanism for COVID-19 infection detection.

COVID-19 is a virus that causes upper respiratory tract and lung infections. The number of cases and deaths increased daily during the pandemic. Once it is vital to diagnose such a disease in a timely manner, the researchers have focused on computer-aided diagnosis systems. Chest X-rays have helped monitor various lung diseases consisting COVID-19. In this study, we proposed a deep transfer learning approach with novel fine-tuning mechanisms to classify COVID-19 from chest X-ray images. We presented one classical and two new fine-tuning mechanisms to increase the model's performance. Two publicly available databases were combined and used for the study, which included 3616 COVID-19 and 1576 normal (healthy) and 4265 pneumonia X-ray images. The models achieved average accuracy rates of 95.62%, 96.10%, and 97.61%, respectively, for 3-class cases with fivefold cross-validation. Numerical results show that the third model reduced 81.92% of the total fine-tuning operations and achieved better results. The proposed approach is quite efficient compared with other state-of-the-art methods of detecting COVID-19.

Prompt Learning for Multi-modal COVID-19 Diagnosis

The outbreak of COVID-19 pandemic has spread rapidly and severely affected all aspects of human lives. Recent researches has shown artificial intelligence and deep learning based approaches have achieved successful results in detecting diseases. How to accurately and quickly detect COVID-19 has always been the core topic of research. In this paper, we propose a novel approach based on prompt learning for COVID-19 diagnosis. Different from the traditional 'pre-training, fine-tuning' paradigm, we propose the prompt-based method that redefine the COVID-19 diagnosis as a masked predict task. Specifically, we adopt an attention mechanism to learn the multi-modal representation of medical image and text, and manually construct a cloze prompt template and a label word set. Selecting the label word corresponding to the maximum probability by pre-training language model. Finally, mapping the prediction results to the disease categories. Experimental results show that our proposed method obtains obvious improvement of 1.2% in terms of Mi-F1 score compared with the state-of-the-art methods. © 2022 IEEE.

Detection of Face Mask Wearing for COVID-19 Protection based on Transfer Learning and Classic CNN Model

In 2020, COVID-19 swept the world. To prevent the spread of the outbreak, it is crucial to ensure that everyone wears a mask during daily travel and in public places. However, relying on human inspection alone is inevitably negligent and there is a potential risk of cross-contamination between people. Automated detection by means of cameras and artificial intelligence becomes a technical solution. By training convolutional neural networks, image recognition can be implemented and image classification can be performed as a solution to the target mask-wearing detection problem. To this end, in this thesis, three typical convolutional neural network architectures, VGG-16, Inception V3, and DenseNet-121, are used as models based on deep learning to investigate the mask-wearing detection problem by using transfer learning ideas. By building six different models and comparing the performance of different typical network architectures on the same dataset using two transfer learning methods, feature extraction and fine-tuning, we can conclude that DenseNet-121 is the typical architecture with the best performance among the three networks, and fine-tuning has better transfer ability than feature extraction in solving the target mask wearing detection problem. © 2022 IEEE.

Prompt Learning for Multi-modal COVID-19 Diagnosis

The outbreak of COVID-19 pandemic has spread rapidly and severely affected all aspects of human lives. Recent researches has shown artificial intelligence and deep learning based approaches have achieved successful results in detecting diseases. How to accurately and quickly detect COVID-19 has always been the core topic of research. In this paper, we propose a novel approach based on prompt learning for COVID-19 diagnosis. Different from the traditional 'pre-training, fine-tuning' paradigm, we propose the prompt-based method that redefine the COVID-19 diagnosis as a masked predict task. Specifically, we adopt an attention mechanism to learn the multi-modal representation of medical image and text, and manually construct a cloze prompt template and a label word set. Selecting the label word corresponding to the maximum probability by pre-training language model. Finally, mapping the prediction results to the disease categories. Experimental results show that our proposed method obtains obvious improvement of 1.2% in terms of Mi-F1 score compared with the state-of-the-art methods. © 2022 IEEE.

Fine-Tuning BERT for Question and Answering Using PubMed Abstract Dataset

The coronavirus, which first originated in China in 2019, spread worldwide and eventually reached a pandemic situation. In the interest of many people, misinformation about the coronavirus has been pouring out on the Internet. We developed a Q&A processing technique by building a dataset based on the PubMed paper for people to easily get the right information. We fine-tuned BioBERT among the BERT models that reached SOTA performance in the biomedical Q&A task. It answered questions about coronavirus with high accuracy. In the future, we will develop our technology that can handle Q&A not only in English but also in multiple languages. This work will contribute to helping people who speak different languages easily obtain correct information amidst confusing data. © 2022 Asia-Pacific of Signal and Information Processing Association (APSIPA).

Detecting Exams Fraud Using Transfer Learning and Fine-Tuning for ResNet50

Online examinations gradually become popular due to Covid 19 pandemic. Environmentally friendly, saving money, and convenient,.. are some of the advantages when taking exams online. Besides its major benefits, online examinations also have some serious adversities, especially integrity and cheating. There are some existing proctoring systems that support anti-cheating, but most of them have a low probability of predicting fraud based on students' gestures and posture. As a result, our article will introduce an online examination called ExamEdu that supports integrity, in which the accuracy of detecting cheating behaviors is 96.09% using transfer learning and fine-tuning for ResNet50 Convolutional Neural Network. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

An Open-Domain QA System for e-Governance

The paper presents an open-domain Question Answering system for Romanian, answering COVID-19 related questions. The QA system pipeline involves automatic question processing, automatic query generation, web searching for the top 10 most relevant documents and answer extraction using a fine-tuned BERT model for Extractive QA, trained on a COVID-19 data set that we have manually created. The paper will present the QA system and its integration with the Romanian language technologies portal RELATE, the COVID-19 data set and different evaluations of the QA performance. © 2022, Institute for Bulgarian Language. All rights reserved.

LDA-CNN: Linear Discriminant Analysis Convolution Neural Network for Periocular Recognition in the Wild

Due to the COVID-19 pandemic, the necessity for a contactless biometric system able to recognize masked faces drew attention to the periocular region as a valuable biometric trait. However, periocular recognition remains challenging for deployments in the wild or in unconstrained environments where images are captured under non-ideal conditions with large variations in illumination, occlusion, pose, and resolution. These variations increase within-class variability and between-class similarity, which degrades the discriminative power of the features extracted from the periocular trait. Despite the remarkable success of convolutional neural network (CNN) training, CNN requires a huge volume of data, which is not available for periocular recognition. In addition, the focus is on reducing the loss between the actual class and the predicted class but not on learning the discriminative features. To address these problems, in this paper we used a pre-trained CNN model as a backbone and introduced an effective deep CNN periocular recognition model, called linear discriminant analysis CNN (LDA-CNN), where an LDA layer was incorporated after the last convolution layer of the backbone model. The LDA layer enforced the model to learn features so that the within-class variation was small, and the between-class separation was large. Finally, a new fully connected (FC) layer with softmax activation was added after the LDA layer, and it was fine-tuned in an end-to-end manner. Our proposed model was extensively evaluated using the following four benchmark unconstrained periocular datasets: UFPR, UBIRIS.v2, VISOB, and UBIPr. The experimental results indicated that LDA-CNN outperformed the state-of-the-art methods for periocular recognition in unconstrained environments. To interpret the performance, we visualized the discriminative power of the features extracted from different layers of the LDA-CNN model using the t-distributed Stochastic Neighboring Embedding (t-SNE) visualization technique. Moreover, we conducted cross-condition experiments (cross-light, cross-sensor, cross-eye, cross-pose, and cross-database) that proved the ability of the proposed model to generalize well to different unconstrained conditions. [ FROM AUTHOR]

Using mobile network data to color epidemic risk maps

In this paper we propose a method for using mobile network data to detect potential COVID-19 hospitalizations and derive corresponding epidemic risk maps. We apply our methods to a dataset from more than 2 million cellphones, collected over the months of March and April in 2020 by a British mobile network provider. The method consists of different algorithms, including detection, filtering, validation and fine-tuning. The approach detected over 2,800 potentially hospitalized individuals, yielding a 98.6% agreement with released public records of patients admitted to NHS hospitals. Analyzing the mobility pattern of these individuals prior to their potential hospitalization, we present a series of risk maps. Compared with census-based maps, our risk maps indicate that the areas of highest risk are not necessarily the most densely populated ones. We also show that the areas of highest risk may change from day to day. Finally, we observe that hospitalized individuals tended to have a higher average mobility than non-hospitalized ones. Overall, we conclude that the rich spatio-temporal information extracted from mobile network data may benefit both the mobile-based technologies and the policies that are being developed against existing and future epidemics. © 2022 ACM.