While browsing new JavaScript libraries to explore, Amanda and I ran into PoseNet, a machine learning model that allows for real-time human pose estimation. We started throwing out ideas, such as dancing games and even yoga poses until we came up with flag signaling, or flag semaphore, which is often used by the US Navy to communicate over long distances. We also thought about the idea of non-verbal communication and came up with the domain for our project, which alludes to the non-verbal and long-distance nature of flag signaling.

What it does

When You Need Personal Space engages with the user by prompting a competitive introduction to flag semaphore. The user must follow the instructions on the page to complete as many flag signals as they can within 10 seconds. The user is prompted with a letter and a corresponding pose that they must recreate. As the user completes more rounds, they will develop muscle memory and think more quickly about each pose, eventually learning at least some form of flag semaphore. Who knows when they will need it?

How we built it

The pose is built with PoseNet using ml5.js library that allows for performance-efficient machine learning done through JavaScript on the web. Animations were created with Adobe After Effects.

Accomplishments that we're proud of

We are proud of finishing this project 7 hours before it was due. As a result, we used the spare time to gamify the web application with a 20-second challenge mode during which the user must go through and try to match poses to letters as they can. Additionally, we created prop flags out of construction paper and pencils that were kindly provided by sponsors at the hackathon for a more realistic demonstration of the app.

What we learned

Aside from strengthening our machine learning application skills, we've learned a thing or two about flag semaphore!

Built With

Share this project: