Family issues and personal experiences.
What it does
This mobile (Android) app calculates the likelihood that a person has cataract(s), corn(s), red eye(s), and/or vitiligo. It also displays historical graphs of each of the values to a user as well as provides resources such as calling for medical help, maps for getting to the nearest medical center, and text/images providing more info on the aforementioned 4 conditions.
How we built it
We used Java to make the Android app and called a REST API to a custom machine learning model that we trained and host in the cloud with AWS. Firebase was used to sync the conditions' data throughout the app.
Challenges we ran into
Finding a lot of training data for the model and its multiple concepts, retrieving data from Firebase, and looking at affected body images to use for our training data.
Accomplishments that we're proud of
Using machine learning for the first time in a hackathon project and finishing the hackathon with a working project.
What we learned
How to use machine learning and integrate it with cloud services.
What's next for MobiDoc
Add detection for more human conditions.