Inspiration
While researching we saw so many images of wounded soldiers, and articles of how many people died because of just a minor cut or burn, as they could not treat or disinfect it properly.
What it does
It takes an image from the user(from the camera roll), and then detects the image. It then takes in the type of wound it is and then displays the right steps to properly treat and disinfect the wound.
How we built it
The first thing we did was to train our machine learning. At first we thought that there could be a framework or API, but unfortunately there were none. Therefore, the only option we have left was to use our own images. We got multiple images from the internet and started adding them into our machine. We continued adding images throughout the contest. After we trained our model, we had to incorporate it into our project. We did this by bundling a local instance of this model into our app. After that, we researched steps to correctly treat the wound. After the user finished all the steps, they could press the back button to identify another wound, or they could just close the application. Now, let's move on to a demo of the app.
Challenges we ran into
Our xcode file became corrupted, and it was hard to increase the accuracy of the AI.
Accomplishments that we're proud of
We were able to make our own machine learning module and incorporate it into our app.
What we learned
We learned how to make a machine learning module, and we learned how to classify pictures.
What's next for InjuryDetector
We want to improve the AI and classify more injuries.
Log in or sign up for Devpost to join the conversation.