Comparison of Fine-tuned CNN Architectures for COVID-19 Infection Diagnosis
Abstract
SARS-CoV-2 (COVID-19) virus spread quickly worldwide affects a variety of industries. The government took preventive steps to control the infection, such as diagnosing the human's lung by taking an X-Ray to see if the lungs were infected with COVID-19 or not. Using several pre-trained Convolutional Neural Network models as the basic model, this research deconstructs the comparison of fine-tuned architecture to identify which pre-trained model delivers the best outcomes in diagnosis by applying machine learning. Comparison is conducted using two scenarios that use batch sizes 64 and 32. Accuracy and f1 score are two evaluation metrics used to justify the model's good performance because the images in the real world, especially for positive classes, are scarce. According to the study, EfficientNetB0 outperforms other pre-trained models, namely ResNet50V2 and Xception, which achieved an accuracy of 0.895 and f1 score of 0.8871.
Downloads
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-ShareAlike International License (CC-BY-SA 4.0) that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
Copyright without Restrictions
The journal allows the author(s) to hold the copyright without restrictions and will retain publishing rights without restrictions.
The submitted papers are assumed to contain no proprietary material unprotected by patent or patent application; responsibility for technical content and for protection of proprietary material rests solely with the author(s) and their organizations and is not the responsibility of the ULTIMATICS or its Editorial Staff. The main (first/corresponding) author is responsible for ensuring that the article has been seen and approved by all the other authors. It is the responsibility of the author to obtain all necessary copyright release permissions for the use of any copyrighted materials in the manuscript prior to the submission.
