LeukoVision: Identifying white blood cells from microscopic images

This project aims to build a deep learning model to identify different types of WBCs. The data that use as a referance for our model is introduced in the following publication: https://doi.org/10.1016/j.cmpb.2019.105020 & https://www.nature.com/articles/s42256-019-0101-9 and the data is avaliable to download by the following link: https://www.kaggle.com/datasets/ahmedali42/dataset-spain-germany

What each folder do?

The folder structure is as follow:

root/ ->contains Readme, gitignore, requirements and all subfolders below
|_src
|_Notebooks
|   |_inceptionv3
|   |_resnet50
|   |_vgg16
|_Streamlit

How to use our models?

To use our model to predict your WBCs, first, we recommend you to use our requirements.txt file to install all the dependant python packages on a new enviroment to make sure that you don't run into any technical problems.

Second, you can download our models (h5 or pth file format) and loaded to identify your images.