This is a binary image classifier. It uses slicing rules, a neural network, and post-processing rules to return a true or false value. See binclass.py for specific options and examples/handwriting/README.md for invocation examples.
Run git submodule update --init to get img2vec and pip install torch torchvision sklearn to get other Python dependencies. On first run, torch will try to fetch the neural networks.