Swifter | Devpost

Skeleton
Object identification

Inspiration

2 people in our group do track and we wanted to create a project that can possibly help us improve our form

What it does

Analyzes a 100m dash by tracking your speed and form every 1.5 seconds by using object identification and pose estimation to create a bounding box and skeleton

How we built it

Python and OpenCV (HTML, Flask were attempted)

Challenges we ran into

Web development was an issue since none of us were fluent with web development, so we could not finish our HTML and implement the website with flask.

Accomplishments that we're proud of

Actually getting the code to work and learning a lot about computer vision. Also, having the stamina to finish.

What we learned

A lot of computer vision and about web dev, specifically about computer vision's capability for analyzing an image. Also, we learned that flask requires time to understand and use.

What's next for Swifter

Maybe making it a personal project or expanding it to be applicable for longer races. Also, we want to be able to compare the skeleton of a user against a professional track runner.