Introducing greenr
greenr is a web app to let users upload an image to classify it as a dandelion or grass.
Click the badge above to run the demo.
greenr is powered by a deep learning model created in FastAI. If you are interested in learning how I created an image dataset from Google Images and trained a dandelions and grass classifier in FastAI, check out this Github repo that links to a notebook outlining my process. FastAI is a deep learning library built on top of PyTorch that makes it extremely to get started with deep learning.
To deploy greenr onto the web, I utilized
- Flask as the back-end to serve the model as an endpoint,
- Flask-CORS to enable Cross Origin Resource Sharing (CORS),
- Docker to containerize the server, and
- Heroku (https://greenr.herokuapp.com/) and Google Cloud Platform (http://34.68.160.231) (for redundancy) to host the container and serve the app.