Posturefy

Inspiration

Our time spent at home during COVID-19 caused us to have bad posture sometimes when sitting at our desks so we wanted to make an application to help keep us healthy and maintaining good posture.

What it does

The application uses machine learning to monitor your body posture and track how long you spend sitting with bad posture and notify you if you have spent too long sitting that way. It also will guide you through stretching routines by monitoring your posture and guiding you into the correct position for the stretch.

How we built it

We built this application using React which uses HTML, CSS, and JavaScript and we also used TensorFlow for the machine learning.

Challenges we ran into

Our largest challenge was producing a machine learning model that accurately analyzes our posture and provides useful feedback that can be presented to the user as well as utilized in powering helpful functions of the application.

Accomplishments that we're proud of

We are extremely proud that we have produced a model that works excellently not only for detecting poor posture but also explaining what is the source of the poor posture so the user can improve and not feel confused as to what they are doing wrong.

What we learned

We learned that it is very important to provide plenty of examples and potential cases for the machine learning model in order for it to be effective in every scenario rather than only a couple.

What's next for Posturefy

The next step for Posturefy is further development on the guided stretching functionality and adding more styling across the application for a better user experience. After this is complete we will have an excellent minimum viable product that we can begin to monetize in a wide variety of ways.

Built With

• react
• tensorflow