Skin Med - Know whats on your skin

Skin Med uses Machine Learning and AI to evaluate and identify any type of skin conditions with the press of a button and seek medical attention accordingly.

Comment

Inspiration

The world is enduring an inevitable fight to the "you-know-why" and we've got nothing to do than just being locked up around walls. People are entitled to a situation where they can't physically seek medical attention. So we developed an app which would ensure the patient's safety during this crisis and simultaneously seek medical attention. Worried about the zit which popped out of nowhere and the ongoing crisis is stopping you from seeking medical assistance? Skin Med have got you covered.

What it does

  1. Classifies the type of Skin Condition of the user from an image using Image processing and Machine Learning.
  2. If the user was diagnosed with an unusual skin condition, he/she can seek medical help by choosing over the wide range of Top Dermatologists in their current location fetched in real-time.
  3. Users could either chat with them or dial the respective dermatology clinic in no time!

It then provides the user with a list of skin care hospitals near the users location and also provides chat feature with the doctor of the selected hospital .The hospitals with highest rating are shown at the first. It also provides contact details of the hospital with a touch of a button to acquire medical assistance.

How we built it

-The front-end was built on Flutter framework.

  • The ML model was built using Python and trained in Tensorflow with Kaggle datasets.

Challenges we ran into

-Training the ML model was a hectic work as it didn't give accurate results at the start.

  • Incorporating the ML model into Flutter
  • Getting list of hospitals using Google Cloud API was a bit difficult.

Accomplishments that we're proud of

-Successfully trained a ML model in a limited time. -Running the ML model completely offline inside the Flutter app.

What we learned

  • Incorporating an exported Machine Learning Model into the app without Internet.

What's next for SkinMed:

Increase the prediction range, and train the model for even more skin conditions.

Built With

Share this project:

Updates