Comparative analysis of Large Image Model and Pre-trained CNN for the Classification of White Blood Cells

White blood cells (WBCs) must be accurately classified in order to be used in medicine for both therapeutic and diagnostic purposes. The performance of two domains that are very efficient in classifying blood cells are examined in this work, with particular emphasis on the models’ robustness and accuracy of classification. The development of Convolutional Neural Networks (CNNs) and, more recently, Large Image Models (LIMs) has drastically changed the landscape of image recognition. This study examines how well these two architectures perform in comparison when used for image-based tasks with uncommon datasets. Although CNNs have proven to be exceptionally powerful in certain fields, LIMs hold the potential to achieve more broad picture understanding. To do this, this research study compares a pre-trained CNN and a cutting-edge LIM. These results clarify the classification of each strategy, showing that CNNs can potentially beat LIMs at jobs requiring specialized expertise. This work adds to the current discussion on model selection in image analysis by highlighting the significance of dataset properties in choosing the best architecture. This research study intends to direct the creation of more efficient image recognition systems for a variety of applications by comprehending the trade-offs between generalization and specialization.