Inspiration
Inspired by the ever-growing advancements in space travel, LiftOff is an AI-powered application designed to address the issue of muscle deterioration in astronauts during their time away from earth.
What it does
LiftOff assesses the accuracy of a user's form when engaging in physical activity, specifically that of astronauts exercising in a zero-gravity environment.
How we built it
We built this in Python using convolutional neural networks to assess whether participants are performing their exercises correctly.
Challenges we ran into
Along with most forms of AI, the lack of sufficient and labeled data was the most significant challenge for us. We needed to gather our own data, requiring time and patience.
Accomplishments that we're proud of
We are proud of our accurate form detector, using a combination of multiple types of AI in the end, providing a 99.98% accurate detection algorithm.
What we learned
In this hackathon, we learned about the challenges associated with AI, like the importance of large amounts of data. We learned how to code a neural network at a low level and integrated this technology to form a complicated algorithm that is able to detect form and improve.
What's next for LiftOff
In the future, we'd like to integrate a broader range of exercises to support specific workout regimes geared towards the user's desired results.
Built With
- ai
- anns
- cnns
- keras
- lstm
- mediapipe
- ml
- python
- tensorflow
Log in or sign up for Devpost to join the conversation.