This is a simple and naive program for deep feature extraction and evaluating face verification performance on the LFW database, implemented by PyCaffe and Python.

I’ve provide a deep caffemodel and trainNet in the model folder, more detail can be found at [1]

For feature extraction, just run the follow command:

 sh runExFea.sh

When feature extraction has done, run:

 sh runROC.sh

for performance evaluation.

The implementation of testROC makes reference to [2], and extraction.py makes reference to the caffe documents.

[1] Deep Learning Face Representation by Joint Identification-Verification

[2] https://github.com/cyh24/Joint-Bayesian