Majority-LCL: Towards Malaria Cell Detection using Label Contrastive Learning and Majority voting Ensembling

Abstract

AbstractThe plasmodium group single-celled parasite that causes malaria is contagious. A female Anopheles mosquito with the infection is most frequently the one spreading it. Over 40% of the world’s population is at danger; 219 million illnesses and around 435,000 deaths were recorded in 2017 alone. Even with many sophisticated evaluation tools, microscopists in resource-constrained settings still struggle to improve the accuracy of diagnosis. Cell picture classification using deep learning avoids making incorrect diagnosis conclusions. This research aims to increase diagnostic accuracy by classifying malarial-infected cells using majority voting ensembling in conjunction with triplet loss-aided label contrastive learning. The outcomes of the experiment demonstrate how well our method works with microscopic cell images in terms of accuracy, precision, recall, and other parameters.