• 

  Logo

• 

  Isometric View

• 

  Landing

• 

  Sign Up

• 

  Data Entry

• 

  Facial Recognition

• 

  Results

• 

  Logo V2

We were really tired coming up with a project.

Introducing Somnus

An app that detects how sleepy you are through facial scanning and calculates the amount of caffeine you need to stay awake and alert.

Tech

To build the app,we used Swift and a Google Auto ML Vision API with our own custom model. To design its interface and prototype, we used Adobe Illustrator, Sketch, and Figma.

View our style guide here: https://goo.gl/h1agXX

Challenges We Faced

• Training our own model to detect drowsiness.

• Integrating Google's Vision API into our app.

• Devising an accurate formula for necessary caffeine intake based on an individual's biometrics and facial scan.

Accomplishments We're Proud Of

• Designing a near-complete user interface and prototyping it.

• Training our model to predict sleepiness fairly accurately.

What's Next

• Continuing to research biometrics that affect alertness in people.

• Working to build our app and publish it.

Built With

• api
• auto-ml
• swift