Eye Biometrics Combined with Periocular and Iris Recognition Using CNN

In this paper, we propose an eye biometric recognition method, which is combined with periocular and iris recognition. The proposed method extracts features using convolutional neural networks (CNNs) dedicated to periocular and iris recognition, respectively. The matching score is calculated by weighting scores from periocular and iris recognition. Through a set of experiments using public datasets, we demonstrate the effectiveness of the proposed method compared with conventional methods.