SpineHealth AI

Inspiration

Posture - the way we sit or stand is the foundation of all the movements our body makes. It is core to how our bodies adapts when subjected under different kinds of stresses. Bad posture results in joints and ligaments being subjected to extra wear and tear further causing increase in the number of physical accidents. It is also proven that lungs become less efficient due to poor body posture. Not only this, our emotional state and sensitivity to pain is also influenced by our posture.

In today's day, people spend time more on their computers than anywhere else which is causing them to look downwards and causing stress to their neck and spine muscles and joints. It has become imperative to create a tool, which like our mothers, can scold us if we are not sitting in the correct position and/or are using our devices continuously for longer periods of time.

What it does

SpineHealth A.I. is an in-browser solution which offers posture correction with the help of object detection and machine learning, offers resources in the form of YouTube videos and articles and provides a compiled report for the posture accuracy and stress levels over the period of a month. It also notifies the user to rest their eyes for 20 seconds after they've been working for 20 minutes thereby following the 20-20 rule.

How we built it

SpineHealth A.I. has been built with ReactJs, MaterialUI, SaSS, Machine Learning and Object Detection.

Accomplishments that we're proud of

Involving machine learning models to not only classify posture of our users as correct or bad but also to display percentage of correctness is an accomplishment our team is immensely proud of.

What we learned

In the duration of this hackathon, our team learned the concepts of MaterialUI and Object Detection.

What's next for SpineHealth AI

SpineHealth AI will soon be available as a browser extension which will enable users to just plug it on and forget about it and get the benefits of posture correction mechanism on any and every website as long as they are online.

Built With

• machine-learning
• materialui
• objectdetection
• react
• sass