I. COVID-19 DETECTION USING DEEP LEARNING MODELS

Abstract

The outbreak of Corona disease, or the so-called Covid-19, has affected the course of human life. Detecting this disease early reduces the risk of spreading the disease. Thus, get rid of this epidemic sooner. In this paper, a system is created that helps to identify and detect Covid-19 disease through X-ray radiation. GoogLeNet, ResNet-101, Inception v3 network, and DAG3Net that are used for comparison purposes. Good results have been obtained in detecting Covid-19 disease, where the DAG3Net produces diagnostic (validation, training, testing and overall) accuracies of (96.15%, 94.34%, 96.75% and 96.58%) respectively, while the GoogLeNet, ResNet-101, and Inception v3 network are produced (98.08%, 100%, 99.59% and 99.72%) respectively.