/domain-specific-BSIF-for-iris-recognition

Domain-Specific Human-InspiredBinarized Statistical Image Features for Iris Recognition

Primary LanguageMATLABGNU General Public License v3.0GPL-3.0

Domain-Specific Human-Inspired Binarized Statistical Image Features for Iris Recognition

Iris image patches used in training and source codes of the method proposed in:

Adam Czajka, Daniel Moreira, Kevin W. Bowyer, Patrick J. Flynn, "Domain-Specific Human-Inspired Binarized Statistical Image Features for Iris Recognition," WACV 2019, Hawaii, 2019

Pre-print available at: https://arxiv.org/abs/1807.05248

Please follow the instructions at https://cvrl.nd.edu/projects/data/ to get copies of these sets.

WACV_2019_Czajka_etal_Stest_GENUINE.csv and WACV_2019_Czajka_etal_Stest_IMPOSTOR.csv files define genuine and impostor matching pairs used to generate Fig. 12 in the paper.

Usage

The example code example.m shows how to use the re-trained BSIF filters in iris recognition.