diff --git a/README.md b/README.md index 98d1ee7..fb7419a 100644 --- a/README.md +++ b/README.md @@ -1,9 +1,8 @@ -# CVNets: A library for training computer vision networks +Enhanced MobileViT for Medical Image Classification (Adapted Using CVNets Library) -CVNets is a computer vision toolkit that allows researchers and engineers to train standard and novel mobile- -and non-mobile computer vision models for variety of tasks, including object classification, object detection, -semantic segmentation, and foundation models (e.g., CLIP). +This project adapts the CVNets computer vision toolkit to train an enhanced version of the MobileViT architecture specifically for medical image classification. While CVNets is a powerful library designed for training both mobile and non-mobile vision models across tasks such as image classification, detection, segmentation, and foundation models (e.g., CLIP), our work customizes it for disease detection using the NIH Malaria Cell Images Dataset. +By integrating the MobileViT framework with domain-specific enhancements—including medical-appropriate data augmentation, multi-scale feature fusion, and attention mechanisms (CBAM)—this project demonstrates how CVNets can be extended beyond general-purpose datasets (e.g., ImageNet) to support high-impact medical imaging applications. The updated system enables efficient and accurate classification suitable for use in clinical settings and on resource-constrained devices. ## Table of contents * [What's new?](#whats-new)