move.io

Leaderboard
Recap of exercise
Profile page
Challenge
Classes
Home
Opencv in action

Inspiration

Through covid we saw a huge increase of virtual learning. We learned how important it is to maximize the little equipment we have in our surroundings. We felt that a lot of people were trying to do home training however, many have had a hard time as motivation was difficult, and lack of education for forms kept them from succeeding.

What it does

Our goal is to provide a service which first educates people on proper form and routines, while also providing incentives to commit to their fitness journey goals. To do so we have 4 main parts of our website. Challenge, Class, Leaderboard, and Profile. Each day, there will be a unique challenge that everyone will be able to participate in. The scores from this challenge will be displayed in the Leaderboard tab. This is to bring out the competitive side in everyone and to encourage users to stay consistent in order to climb the rank. We also have more consistent exercises and education in the Classes tab. Users can pick out what exercises they would like to do and their exercise history, as well as their daily streaks and more can be found in the Profile Tab. These were designed to encourage users to come back to the website and stay consistent in their fitness journey at home. We used opencv in order to provide real-time feedback on the user’s forms by applying algorithms to detect and give feedback to the user about their form. This way, they can efficiently and safely exercise while staying motivated and confident about their exercises.

How we built it

To Power the website we used Flask as the backend, where we used OpenCV to help us analyze the user’s movements. We used html/css/js for the frontend. We used REST API to connect the two. To create a user friendly interface, we started brainstorming the website design with sketches. From there, we went into Figma to iterate and develop higher fidelity prototypes. Once we had a good idea of what we wanted the final product to be, We used html, css, and javascript, to create reactive pages that allow for a simple, yet exciting experience to navigate through the website.

Challenges we ran into

Tuning the algorithm to make it consistent in different exercises, users, lighting, and angles was very difficult. This was because even a little difference in the environment could dramatically impact how the camera picks up movement.

Accomplishments that we're proud of

We were really proud that we were able to complete the project with the many of the functionalities we envisioned when we started.

What we learned

We learned how to make tough crunch time decisions and to always be flexible. In such a short time period, we had to consistently change our plans on how something is implemented and modified.