Dermetric is a machine learning-based diagnostic tool for skin diseases made by the team of Cashel Fitzgerald, David Pekar, James Liu, Carine Harb, Aniket Gupta, Advik Sundar, Arjun Kumar, Minjae Cho, Pranav Tonpe, Arnav Chopra, and Lakshay Nagpal.
The tool is powered by computer vision and a mobile application. So far, Dermetric has implemented a TensorFlow-based machine learning model for classification of various skin disorders, along with backend functionalities like model integration and image input, complemented by a designed Figma wireframe for the user interface.
cashel.dev/Dermetric
- Hosted on AWS
- Running Back-End API on a Flask web server
- Completed Front-End via React Native (currently compiles to IOS, Android, and a web app)
- Completed a CNN classification model in TensorFlow (ML)
- Increase accuracy of the skin disease diagnostic model to at least a 90% accuracy rate
- Develop a splash screen for the mobile app
- Add a treatment recommendation page in the app that shows users the most nearby and suitable clinic given their condition
- Deploy the app on Google Play Store and Apple App Store