PoxNet22: A Fine-Tuned Model for the Classification of Monkeypox Disease Using Transfer Learning

Abstract

Officials in the field of public health are concerned about a new monkeypox outbreak, even though the world is now experiencing an epidemic of COVID-19. Similar to variola, cowpox, and vaccinia, an orthopoxvirus with two double-stranded strands causes monkeypox. The present pandemic has been propagated sexually on a massive scale, particularly among individuals who identify as gay or bisexual. In this instance, the speed with which monkeypox was diagnosed is the most important aspect. It is possible that the technology of machine learning could be of significant assistance in accurately diagnosing the monkeypox sickness before it can spread to more people. This study aims to determine a solution to the problem by developing a model for the diagnosis of monkeypox through machine learning and image processing methods. To accomplish this, data augmentation approaches have been applied to avoid the chances of the model’s overfitting. Then, the transfer-learning strategy was utilized to apply the preprocessed dataset to a total of six different Deep Learning (DL) models. The model with the best precision, recall, and accuracy performance matrices was selected after those three metrics were compared to one another. A model called “PoxNet22” has been proposed by performing fine-tuning the model that has performed the best. PoxNet22 outperforms other methods in its classification of monkeypox, which it does with 100% precision, recall, and accuracy. The outcomes of this study will prove to be extremely helpful to clinicians in the process of classifying and diagnosing monkeypox sickness.

Keywords

• Monkeypox
• data augmentation
• transfer learning
• classification
• PoxNet22