Breadorcat

A web service solving rocket science task - distinguishing cats and bread

Model

Download model here

prepare_data.py collects data from the folder ./source_data, which consists of 4 folders containing unarchived open source datasets: Tiny ImageNet, Dogs vs. Cats, Food-11 and iFood-2019. Note: hierarchy ./data/{train, test}/{bread, cat, other} must be preapred in advance.

Data collected with the script above can be downloaded here

Backend

To start backend:

virtualenv .env --python=python3.7
. .env/bin/activate
pip install -r requirements.txt
python main.py -p 8080

or

virtualenv .env --python=python3.7
. .env/bin/activate
pip install -r requirements.txt
make runserver