Taking a photo of your skin is as trivial as taking a photo to anything else; with that fact, we took advantage of the accessibility of machine learning algorithms and public images of skin diseases to train our model.

What it does

Our web application receives an image from the user, then it uses our trained deep learning algorithm to display a probability of what kind of disease you may have.

How we built it

We used a responsive web application for the front-end where you upload a photo of your skin and then it is send to a server that has a trained deep learning model to classify it.

Challenges we ran into

We had to create a small script to retrieve images of skin diseases from website since it did not have a download button

What we learned

  • We learned responsive style for websites
  • We learned to use heroku to easily deploy the project
  • Pytorch

What's next for Skin Doctor

Create a native application either for android, IOS, or both in order to improve user experience.

Share this project: