dermaCare

• 

  The home page that has the feature to upload an image to the web app

• 

  The prediction page that displays the mole's characteristic

What it does

The webapp consists of an ML model that predicts if the image uploaded to it could be cancerous or not.

How we built it

We first made the ML model using Tensorflow with Jupyter Notebook and then wrote a python script that uses the model and returns the prediction. We then integrated the script with the flask server and used CSS styling to decorate the webpage.

Challenges we ran into

1. Building the ML model was a daunting task.
2. Linking the ML model to a separate python script was difficult.
3. We ran into a lot of errors by the time we had the flask backend up and running

What we learned

1. The process of image recognition using Convoluted Neural Networks.
2. This our first time working with the Flask framework and we found it to be quite interesting
3. Worked with some cool python libraries that made our work a lot more easier ## What's next for dermaCare
4. As of now, the model has a 70% accuracy. We are planning to train the ML model with an even larger data set to increase its accuracy.
5. We are planning on integrating a front end framework like react soon enough to the project

Built With

• flask
• python
• tensorflow