Sparse Support Faces (SSF) Code Project

A Super-sparse biometric verification system drastically reduces the number of reference templates per client, saving memory and computational resources during client verification and identification, without affecting recognition accuracy.

To foster reproducible research, we decided to release a revised version of the code used in our paper "Sparse Support Faces". It enables reproducing some of the results reported in the paper, and contains the implementation of the proposed algorithm for super-sparse learning of face templates in biometric identity verification settings.

The code is released under GNU GPL 3.0, and it can be downloaded here

Please refer to the README file for instructions on how to install the required libraries and use the code.

If you use this code for research or similar purposes, please cite the reference paper:

B. Biggio, M. Melis, G. Fumera, F. Roli, "Sparse Support Faces", in Int'l Conf. on Biometrics (ICB), In press.

Your feedback is also appreciated. Please contact the authors for any enquiries.