Malar-ai is a Mobile Application used to detect probable malaria infection using blood smear slides.
Powered by ResNet-50, cell-phone microscope and a staining kit, this setup is capable of performing at an accuracy of >95%, hence highly useful in remote and inaccessible areas.
We aim to enhance the ease of Malaria testing in areas where the availability of testing equipments is not adequate. Using this mobile app health care worker can capture the photo of a blood smear slide of a person, and predict the possibility of Malaria.
You can have a look at the source code of the developed machine learning model used by this application here and this repo contains the source code of the android app and potentially a lens .
This project is maintained by the students Club of Sustainability & Innovation, Indian Institute of Technology(IIT-BHU), Varanasi.