PocketDerm

• 

  Recommending skincare products based on scan

• 

  How the recommendation system works

• 

  Skin disease detection scan using roboflow api

• 

  Our personalized chatbot

Inspiration

The skin is the largest organ in the body, and one that usually goes unnoticed by millions of people as a part of their overall health. There are also tons of different types of skin, so no skincare product can be one size fits all. One product might be the saving grace for someone, while giving someone else inflammation. There is overall a lot of confusion over skin health and this is what PocketDerm solves.

What it does

The website's main feature is the skincare product recommendation system, which scans the user's face and what type of products they want, and from that it gives them a list of recommended products based on what others with the same skin thought of the product. Another feature is skin disease detection through another scan of the user's skin. The last feature is an embedded chatbot which can assist user's through attaining better skin health.

How we built it

I built this using pytorch and roboflow's inference api for the scannings, django for the backend and html + css + bootstrap for the frontend.

Challenges we ran into

I ran into problems on figuring out how I would make the recommendation system.

What we learned

I learned more on how to use the roboflow inference api.

What's next for PocketDerm

I wish to make a page for tutorials and a page to help people stay on top of their skincare routine.

Built With

• bootstrap
• css
• django
• html5
• pytorch
• roboflow