Inspiration
We wanted to get better at sports, but we don't have that much time to perfect our moves.
What it does
Compares your athletic abilities to other users by building skeletons of both people and showing you where you can improve.
Uses ML to compare your form to a professional's form.
Tells you improvements.
How I built it
We used OpenPose to train a dataset we found online and added our own members to train for certain skills. Backend was made in python which takes the skeletons and compares them to our database of trained models to see how you preform. The skeleton for both videos are combined side by side in a video and sent to our react frontend.
Challenges I ran into
Having multiple libraries out of date and having to compare skeletons.
Accomplishments that I'm proud of
What I learned
What's next for trainYou
Built With
- google-cloud
- python
- react
- tensorflow

Log in or sign up for Devpost to join the conversation.