COVID-19 Detector Using Deep Learning

Abstract

This paper attempt to overcome the existing Covid 19 detection challenge, which aims to predict whether the tested person is a covid positive or covid negative. This research work has utilized the “Covid chest X-Ray” images dataset and the CT scan images dataset of Covid affected people and healthy people from the Kaggle website. Further, the proposed research work has utilized a couple CNN models on the collected dataset to see if the input image was Covid positive or negative. This research study will construct four CNN architectures as a group: ResNet-50, Inception-v3, and Xception. These models will be trained by using chest X-Ray and CT scan images, and then a Web API will be developed by using Flask so that the users may interact within themselves via a website. And the user may self-identify whether he/she is Covid positive or negative by uploading any chest X-Ray or CT scan image of the individual.