This is a minimal package that can instantiate and extract features with any CNN model of the family described in [1]. Its purpose is mainly for the reproduction of that and later results. If you want to use it as a simple feature extractor based on the model HT-L3-1st of [1], simply install it with
python setup.py install
and refer to the example code in
examples/extract_features.py
Otherwise, if you want to hack and extend the package, go with:
python setup.py develop
In both cases, check if everything is running smoothly with:
python setup.py nosetests
[1] Nicolas Pinto and David D. Cox, "Beyond Simple Features: A Large-Scale Feature Search Approach to Unconstrained Face Recognition," in IEEE Automatic Face and Gesture Recognition, 2011.