CPR Now is an iOS app that incorporates Stochastic Gradient Descent to interpret speech and decide on the best course of action to administer CPR to another person who may need it. It uses Natural Language Processing to interpret any scenario given by the person conducting CPR. The app gives clear instructions to the user, and has a pulsing UI for the user to sync his/her compression rate. Each step is clearly shown with a graphic image and text to assist the person administering CPR. Audio prompts from the app also help the user understand what to do next. Twilio API is used to send pre-written messages to emergency contacts in case the victim is alone or the CPR conductor requires additional assistance.
Challenges we ran into
As high schoolers, swift was unfamiliar to many of us so we had to learn how to incorporate a visually appealing UI and handle complex control flow events within in the app.
Accomplishments that we're proud of
CPRnow can potentially save lives in the future by not only giving everyone easy access to personalized CPR instructions but also to instant emergency messaging and communication courtesy of Twilio.
How we built it
- Converted spoken word into text and extrapolated keywords with natural language processing.
- Stochastic gradient descent allows for prediction of the CPR "step" from audio input.
- Created a modern UI, and displayed step-by-step instructions for CPR accompanied by instructions.
- Created a pulsing circle to aid in chest pumps.
- Hosted our model and Twilio functions in a Flask REST Api on a server.
What we learned
- How to effectively use swift when creating iOS apps and how to implement more advanced machine learning algorithms from python into mobile apps.
- Managing control flow in complex apps, and balancing out the need for a nice UI and clean programming.
What's Next?
Improve and fully connect UI Multi-language support Siri Integration Use custom domain to host server and functions ** **The world!
Log in or sign up for Devpost to join the conversation.