An optimized transfer learning-based approach for automatic diagnosis of COVID-19 from chest x-ray images.

Accurate and fast detection of COVID-19 patients is crucial to control this pandemic. Due to the scarcity of COVID-19 testing kits, especially in developing countries, there is a crucial need to rely on alternative diagnosis methods. Deep learning architectures built on image modalities can speed up the COVID-19 pneumonia classification from other types of pneumonia.