Building an App to help people not touching their faces and get infected.
What it does
Notify smartwatch users when they're about to touch their faces so they get reminded not to.
How we built it
We built it with ML technologies and used CoreML in the end to build a Prototype for the Apple Watch.
Challenges we ran into
Gathering Training Data, putting a ML model on the Apple Watch, running an Apple Watch App in the Background.
Accomplishments that we're proud of
We are proud that we were able to develop a PoC in such a short time. It does not work perfectly yet, but it shows what is possible.
What we learned
ML is hard work. Data gathering is very difficult.
What's next for Guardy - The Face Touching Detector
- Implement Apps for smart watches
- Train guardy with more data to detect motions more accurately
- Constructing a user friendly app environment
- Social features like sharing and gamification
- Ask the users to start Guardy as son as they leave the house through location data
Try It out
android, coreml, garmin, machine-learning, monkeyc, python, r, swift, watchos