This is an artwork classifier, which classifies paintings of more than 200 artists such as Claude Monet and Pablo Picasso. A running demo is deployed at Heroku.
The classifier is based on a deep neural network (DNN) implemented and trained with tensorflow. The model architecture is constituted by a multi-layer convolutional neural network, which is followed by a fully connected layer and completed with a softmax readout layer.
Training was done based on a cross-entropy error function.
$ virtualenv -p python2.7 venv
$ source venv/bin/activate
$ pip install -r requirements.txt
$ gunicorn main:app --log-file=- --timeout=600 --preload
$ python preprocess.py
$ python train.py