moleCare - CNN project

moleCare is an application that uses Convolutional Neural Networks (CNN) to classify images of skin moles as malignant or benign.

Some highlights:

• We applied transfer learning and used the pre-trained MobileNet v2 model to train our data. Additional layers were added to fine tune our model.
• The API was developed with Flask framework.
• The application was built in a Docker, and the resulting Docker image was then deployed to Heroku.