Classification of Cells Infected by Malaria with Vision Transformers Models

This project develops a system for malaria classification using the Vision Transformer architecture on cellular tissue microscopy images. The implementation aims to assist in medical diagnosis by providing efficient and accurate information, with the potential to contribute to the control of endemic diseases. The system utilizes models from the vit-pytorch library, including the Vision Transformer, Patch Merger, and Vision Transformer for Small Datasets. The Vision Transformer for Small Datasets achieved the highest accuracy of 98.97 for an image size of [150,150].