Inspiration:
We were interested in how machine learning has recently taken the world by storm, and we wanted to create a web app that can help people somehow. We noticed that many of our group members didn't know how to deal with things they noticed about their skin.
What it does
We decided to make a free version of an AI that analyzes the skin from a picture.
How we built it
We used TensorFlow along with Streamlit to create a web app.
Challenges we ran into
Ideas; we spent more time on that phase than we would like to admit. :P We initially wanted to make a fancy UI but soon realized it was too ambitious, so we had to learn to scale down our scope and have a product that worked. It was also hard to trick the AI that I (the funny man) had cancer, because it was really well trained to the point that it recognized itty-bitty reflections and began to point out that my "cancer" was fake (obvious opp).
Accomplishments that we're proud of
We were able to scale down our idea while retaining its technical complexities. We also ended up going beyond our original tiny plan.
What we learned
We learned how to load models and create our web apps from them.
What's next for Dermalyze
We truly enjoyed working on this project and would like to continue developing our own model by making the training and pre-processing steps more optimal. We intend to keep this app as accessible as possible because we found that many of the AIs out there are locked behind a paywall. By keeping our app free, we can increase awareness about existing solutions. To maintain our project's sustainability, we can look into using affiliate links, such as Amazon, to provide users with links to buy products instead of just stating them, allowing us to pocket some money to invest in keeping a server online.
Log in or sign up for Devpost to join the conversation.