Inspiration
Many people spend an increasing amount of time doing desk jobs and have poor seated posture that could lead to many health issues like back pain, headaches, and spine curvature. We aim to develop a unique application that will support people to improve their posture by using the phone's camera to monitor the user while they are seated and notify them whenever they get into a bad posture. Although there are other posture correction applications on the google play store or apple's application store, almost all of them involve some expensive tracker placed on the body and does not seem to use the vast capabilities of Machine learning and Computer vision. The spine is one of the essential parts of the body. Without it, we cannot keep ourselves upright or stand up. It gives us body structure and support. It allows you to move freely and is designed to protect the spinal cord. Without the spinal cord, we cannot move even an inch; hence keeping the spine healthy is vital if you want to live a happy life. This is why we want to make a profound impact on everyone's lives by developing an innovative product that will be immensely beneficial. This was our class project in which we developed the posture detection, we further extended this application by adding a user statistics activity that tracks how many times a user gets into bad posture. We also added a bad posture gallery consisting of the highlights so that the user can review and improve their posture later on.
What it does
It helps users analyze their posture and gives recommendations on their performance. The app constantly monitors the person's stance at regular intervals and tries to remind the user to correct their posture with alarms of gradually increasing volume to attract the user's attention. Throughout each session, the user can see the highlight snippets captured while he was not in posture so that they can improve it during the upcoming session. The application also allows users to talk to professionals and get medical help for incorrect posture. The reason for this is that as soon as the app detects that a posture is incorrect, its not safe for the passengers to travel to a doctor and sitting with bad posture in the car. On the other hand, the cost of an ambulance is really high, so through our app the doctor can directly access the metrics of the user's performance and defects to give a temporary solution to ensure that the person can travel properly to the hospital and cure it on a long term basis.
How we built it
The android application was built using Java and Firebase. We also utilized python, scikit-learn and Tensorflow. In order to perform an analysis of the user's posture,we are using OpenCV and Mediapipe libraries to find the coordinates of various landmarks like the shoulders, back, neck, etc of the user ultimately which are passed to a custom neural network learning model that classifies the coordinates as good or bad posture.
Challenges we ran into
There was a huge learning curve for implementing neural networks, and we were able to only do it by the combined and collaborative effort of all our team members by effectively solving all the error while debugging.
It was really hard in getting that the neural network to work in android.
Accomplishments that we're proud of
We were able to learn the fundamentals of neural networks along the way. We were able to supplement each other while working on the project and utilize the time effectively.
What we learned
We learnt how to use neural networks in conjunction with android. We also learnt how to implement firebase and integrating our back-end.
What's next for SimpleStance
Next, we will be working on making it a better functioning app by making the UI/UX more advanced and we will continue to use the domain and theme as it resonates with not only the purpose of the application but also has a touch of eastern culture.
Log in or sign up for Devpost to join the conversation.