Abstract:

More than a million people have been infected with the newly discovered coronavirus (COVID-19), and over 50,000 have lost their lives. The World Health Organization has designated this virus an epidemic. The development of pneumonia from a COVID-19 infection is detectable with a lung X-ray and requires medical attention. In this paper, we argue for a fully automatic method of detecting COVID-19 pollution in lung X-rays. The research files include 194 X-ray images, 194 of which were taken from patients who had been diagnosed with coronavirus, and 194 from healthy patients. With so few openly accessible images of people with COVID-19, we are forced to rely on the concept of transfer learning to complete this assignment. To retrieve features from X-ray pictures, we modify the behaviour of convolutional neural networks (CNNs) with custom designs that have been trained on ImageNet.

Description:

.