ABSTRACT
In recent months, the fight against COVID-19 has grown into one of the most actively pursued anti-toxin treatment strategies worldwide. Correct medical reasoning and a swift response are essential to preventing the COVID-19 epidemic from taking an unexpected turn. Corona virus can be detected via RT-PCR, although chest X-ray techniques have been more effective and helpful in detecting the virus's effects. With an increasing number of people being detected with COVID and a larger number of X-rays being taken, it is now viable to use transfer learning to categorise the X-ray results. Covid19, bacterial pneumonia, and normal incident X-ray datasets have been combined to develop an automatic method for detecting the disease. Specifically, the objective of this study is to achieve better image classification results over state-of-the-art models like the Convolutional Neural Network (CNN) that were developed recently. The data sets were collected from freely accessible online medical sources. The results shows that significant biomarkers associated with Covid-19 illness can be identified using a combination of Transfer Learning and X-ray imaging. In our experiments, we found that the best accuracy was achieved using a combination of VGG16, Resnet50, and a Convolutional Layer, with respective values of 96.78, 98.66%, and 96.46 %. X-rays' potential for utility in diagnosis has grown as the failure rates of older, more established analytical methods have grown alarmingly high. Copyright © 2022, Anka Publishers. All rights reserved.