COVID-19 Semantic Search Engine Using Sentence-Transformer Models

With the onset of COVID-19, enormous research papers are being published with unprecedented information. It is impractical for the stake holders in medical domain to keep in pace with the new knowledge being generated by reading the entire research papers and articles in order to keep pace with new information. In this work, a semantic search engine is proposed that utilises different sentence transformer models such as BERT, DistilBERT, RoBERTa, ALBERT and DistilRoBERTa for semantic retrieval of information based on the query provided by the user. These models begin by collecting COVID-19-related research papers and are used as an input to the pre-trained sentence transformer models. The collected research papers are then converted into embedded paragraphs, and the input query is sent to the same model, which in turn delivers the embedded query. The model uses cosine similarity to compare both embedded paragraphs and the embedded query. Consequently, it returns the top K most similar paragraphs, together with their paper ID, title, , and summary. The bidirectional nature of the sentence transformer models allows them to read text sequences from both directions, making the text sequence more meaningful. Using these models, COVID-19 semantic search engine has been developed and deployed for efficient query processing. The similarity score for each model was computed by averaging the top 100 query scores. As a result, the RoBERTa model is faster, generates a higher score of similarity, and consumes less runtime. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Learning from YouTube Videos using Drona Extension

There has been a boom in the use of online resources for education since the onset of the COVID-19 pandemic. Educational videos are an important part of these online resources and are very helpful to students. However, there are various issues related to the use of educational videos such as losing focus, lack of feedback to students, and taking notes manually while watching. The primary objective of this work is to design and develop an intelligent browser action chrome extension that generates smart notes and quizzes from educational YouTube videos. The extension will serve anyone learning from YouTube by providing a dynamic platform to grasp the content and assess themselves. © 2021 IEEE.

Real time Face Detection and Optimal Face Mapping for Online Classes

The main objective of this paper is to provide a web-based tool for identifying faces in a real-time environment, such as Online Classes. Face recognition in real-time is now a fascinating field with an ever-increasing challenge such as light variations, occlusion, variation in facial expressions, etc. During the current pandemic scenario of COVID-19, the demand for online classrooms has rapidly increased. This has escalated the need for a real-time, economic, simple, and convenient way to track the attendance of the students in a live classroom. This paper addresses the aforementioned issue by proposing a real-time online attendance system. Two alternative face recognition algorithms are perceived in order to develop the tool for real-time face detection and recognition with improved accuracy. The algorithms adopted are Local Binary Pattern Histogram(LBPH) and Convolutional Neural Network (CNN) for face recognition as well as Haar cascade classifier with boosting for face detection. Experimental results show that CNN with an accuracy of 95% is better in this context than LBPH that yields an accuracy of 78%. © 2022 Institute of Physics Publishing. All rights reserved.

Perceiving Machine Learning Algorithms to Analyze COVID-19 Radiographs

Coronavirus disease, also referred to as COVID-19, is a contagious illness generated by a respiratory virus. There has been an exponential increase with the amount of patients affected with COVID-19 that has put an exceptional burden on the medical care frameworks across the world. Analysis of COVID-19 disease from the images of Chest X-ray may help isolate high-risk patients, while test results are anticipated upon. With most X-ray frameworks currently digitized, there is no transportation time required for the samples, hence making it easier for the health care workers to analyze it. In this work, we demonstrate the potential of ResNet, which is a CNN, to diagnose Chest X-ray images. These images can be classified into Normal, COVID, or Viral Pneumonia efficiently using ResNet. As a result, the probability of detecting patients with COVID-19 is maximized through higher accuracy. Empirical analysis exhibits that the proposed neural network strategy is better than Support Vector Machine, Naive Bayes algorithm, Logistic Regression, and k-NN. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.