Inspiration
Fitness plays a crucial role in most of our lives in regard to mental and physical health. Our passion for fitness and leading a healthy lifestyle played a massive role in choosing this topic, especially with regards to how it affects the current generation and how it may affect future generations. In addition, many studies have concluded that our upcoming generation is seemingly becoming lazier by the second. One of us lives with younger siblings and can vouch for the fact that their regard for health and fitness is far below what it should be. However, with that said, it's easy to see that their interest in video games has started to peak. Video games in general may not be seen as beneficial, however, it has the ability to captivate individuals from around the world. We've decided to utilise that strength to benefit users by encouraging them to pay more attention to their physical health.
What it does
LevelUp is a software that is able to analyze users as they exercise in order to provide visual feedback to ensure they are able to maintain good form and develop a healthy workout routine. We provide two different game modes, namely Progress and Infinite. Progress provides users with a chance to hit their target while showing them a visual representation of their workout progress.
How we built it
The project, which was built in python, mainly utilizes three different frameworks, mediapipe, openCV, and pyGame. The frontend of the software was designed mainly using pyGame. This provides users with an intuitive layout and easy-to-use platform. In regards to the backend, our software's technical features are spotlighted by the use of the combination of OpenCV and Mediapipe. Mediapipe provided a crucial role in our project with its efficiency in analyzing human poses. OpenCV provided the means to convert the mediapipe landmarks into useful pieces of data which were useful in various technical features such as angle detection.
Challenges we ran into
Throughout this project, our main challenges revolved around the technical features. One challenge that we faced was the camera's orientation as in specific moments, finding the angle between certain areas would be rather difficult with the camera receiving little visibility.
Accomplishments that we're proud of
We felt extremely proud to be able to learn different frameworks and be able to work along with one another in an extremely tiring and difficult 2 days.
What we learned
One fascinating topic we learned was to work collaboratively using GitHub as an intermediary and the basics of Machine Learning.
What's next for Level Up
In order to Level Up our software, one of the key missing factors in our current project is the need for a leaderboard (both personal and worldwide), which tracks records of each person's daily workout routine. This would be easy to implement using frameworks and databases like MongoDB or CockroachDB to extend our software on a global basis. Moreover, a key component of most people's workouts is music. As such, we see possibilities to utilize the spotipy API to include a large data set of music to provide various options for our client to meet their emotional needs. Finally, the last big change would relate to increasing the current number of exercises our software can host. At the moment, we have successfully developed 3 popular exercises that target the entire body. Likely, we will increase the number of exercises and incorporate object detection to provide workouts with related equipment.
Log in or sign up for Devpost to join the conversation.