What would you say if I told you there is an app on the market that could tell whether your pizza has pineapple on it?
Convolutional neural network built with Keras. It uses the PizzaGAN labeled dataset (see the scripts in dataset/ directory). For better accuracy add more
images of pineapple pizzas!
I got the idea from the Silicon Valley not hotdog app.
- Install tensorflow,
pip install -requirements.txtand whatever it wants. - Run
dataset/download_PizzaGAN.shto download their dataset (~2.8GB + unzipped). - Tweak the settings in the dataset/categorize.py script and run it.
- Download more images of pineapple pizzas in
dataset/training_set/pineapplefor better accuracy. - Run the
classifier.py/classifier.ipynb(for example with jupyter or IDE) and start training.
- Install the requirements (same as above)
- Start with
cd server && flash run - Navigate to the development server at http://127.0.0.1:5000/
Disclaimer: I had no idea what I was doing. I am not responsible for any pineapple pizza related injuries.