A simple implementation of the Google Quickdraw Dataset. Run main.py and draw. Press space to get a prediction and right click to reset.
Using a subset of a subset of the 50 million drawings, I trained a convolutional neural network to recognize all 345 classes (It's not great at it though).
I wanted to learn how to use tensorflow and keras, and I thought this would be a fun way to do it. I've been kinda absent from Github for a little bit and thought this was a fun little experiment to show why.