Worried about your Rash? Can't wait to visit a doctor or are just a wee-bit curious? What if it's just bite are you but you can't help be paranoid what if it's a malignant beginning to cancer? Here's your solution
Collecting the data
We used the official HAM10000 dataset. 3 major caner rashes were choosen for MVP
(Melanoma, Benign Keratosis and melanocytic Nevi).
Detection Software
My used CNN, implemented Yolo-V3 and trained a model to diagnose the type of Rash
APP connectivy
Uploaded model to firebase, created API and connected it a simple android app, with end points being the image and the diagnosed ailment.
Result
Working Under Dr. Sharnil Pandya, Professor at Symbioses my team backed the 3rd prize at the startup expo
Technologies used and skills learned were:
- Python3
- Tensorflow
- Cloud connectivity
- Firebase
- YoloV3
- Communication skills
- Study on Image preparation for training