Inspiration
What it does
Gives out the energy consumed during activity without any fit band.
How I built it
we used Posenet to understand the body posture and how the movement is happening with respect to time and based on that, we have built a way to understand the energy consumed during the activity.
Challenges I ran into
It is during training, we were unable to find datasets, for which we wanted to predict more than 80 points on the body. but, somehow, we were able to implement it with a minimal set of points. We did face a few problems during the implementation of the same on IOS and Android
Accomplishments that I'm proud of
The ability to make these models light, so that they worked near-realtime on edge devices
What I learned
What's next for FEnergy
We want to add features like pose correction during the exercise
Built With
- android
- ios
- javascript
- python
- swift
- tensorflow
Log in or sign up for Devpost to join the conversation.