COVID-DETECT: A Deep Learning Based Approach to Accelerate COVID-19 Detection

June 2021
Engineering and Numerical Tools
Communications avec actes dans un congrès international
Auteurs : Samir Ouchani (LINEACT)
Conférence : Intelligent and Autonumuous Systems, 18 June 2021

After two years of COVID-19 first infection and its speedy propagation, death and infection cases are till exponentially increasing. Unfortunately, during this a non-fully controlled situation, we noticed that the existing solutions for COVID-19 detection based on chest X-ray were not reliable enough in relation to the number of infected patients and the severity of the outbreak. To handle this issue by increasing the reliability and the efficiency of COVID-19 detection, we therefore deploy and compare the results of a set of reconfigurable classification approaches and deep learning techniques. Indeed, we have achieved a score of up to 99% AQ1 accuracy with a dataset of 15,000 X-ray images, which makes the selected detection technique, deep learning, more reliable and effective.