Affordable Access

deepdyve-link
Publisher Website

Investigation and Implementation of Convolutional Neural Networks with Transfer Learning for Detection of Covid-19

Authors
  • Raghavendra, Ch
  • Phani Sriram, S
  • Sravan Kumar, V
  • Sudheksha, K
  • Bhavya Sri, P
Type
Published Article
Journal
Journal of Physics Conference Series
Publisher
IOP Publishing
Publication Date
Sep 01, 2022
Volume
2335
Issue
1
Identifiers
DOI: 10.1088/1742-6596/2335/1/012023
Source
ioppublishing
Keywords
Disciplines
  • Paper
License
Unknown

Abstract

This paper aims to present a robust model which uses Artificial Intelligence in the rapid and accurate detection of Covid-19. The Proposed Project uses Deep Learning techniques for the detection of Covid-19 as well as Pneumonia with the help of Digitalized Chest X-rays. By Applying Pre-Trained Networks which are also known as Convolutional Neural Networks, this can be made possible. Also, with the help of Transfer Learning Techniques the Neural Networks can be trained and validated faster and better thus providing a significantly higher chance in the correct and accurate detection. The CNNs taken into consideration are VGG-19 and ResNet-50 with the former providing an accuracy of over 95% and the latter with an accuracy of 92% proving that such a high accuracy computer-aided diagnostic tool can be used at a time like this. Also, further improvements in the future with advancements in technology can provide even astonishing results.

Report this publication

Statistics

Seen <100 times