The International Skin Imaging Collaboration: Melanoma Project is an academia and industry partnership designed to facilitate the application of digital skin imaging to help reduce melanoma mortality. When recognized and treated in its earliest stages, melanoma is readily curable. Based on the ISIC Detection Dataset with over 23.000 images of skin lesions, labeled as 'benign' or 'malignant', the web application provides a Hands on API for uploading and classifiying and image easily.
Included in the WebApp is a SignUp and Login functionality as well as a database provided for user account information all based onto firebase. In addition to it a paywall is added with two options: Paypal and Stripe. So all potential users should be able to handle a payment for a fast and accurate diagnosis.
- Clone this repo
- Install requirements
- Run the script
- Check http://localhost:5000
- Login with a test user account: E-Mail: test@test.com | PW: 12345678
- Done! 🎉
git clone https://github.com/JanMarcelKezmann/python-flask-firebase-web-application.gitpip install -r requirements.txtMake sure you have the following installed:
- tensorflow
- keras
- flask
- pillow
- h5py
- gevent
- flask_dropzone
- pyrebase
- numpy
- stripe
- paypalrestsdk
- werkzeug
Python 3.6 is supported and tested.
python app.pyOpen http://localhost:5000 or http://127.0.0.1:5000 and have fun. 😃
Training the model with a on the imagenet dataset pretrained InceptionV3 model gives the follwing scores:
- Accuracy: 0.831
- F1 Score: 0.8456
- Precision Score: 0.925
- Recall Score: 0.7786
The Confusion Matrix of the Model evaluated on 2000 images
Pick the login E-Mail and type in the credentials shown in the picture below
For testing Paypal payment a developer account is neccessary. Here is the test-e-mail and test-password I used:
E-Mail: sb-ychg6329905@personal.example.com | PW: J[0k$44M