Personal Authentication of Iris and Periocular Recognition using XGBoost
Daisuke Uenoyama, Hiroshi Yoshiura, Masatsugu Ichino · 2019
Iris authentication is attracting increasing attention due to its high accuracy. However, it imposes a psychological burden because the person to be authenticated must closely approach the camera in order for it to capture a high-quality image of the person's iris region. One way to reduce the burden is to combine iris authentication with periocular authentication. Following this approach, we focused on increasing the accuracy of iris authentication at a distance by using more periocular features and the XGBoost algorithm to fuse the scores. Test results show that our proposed method is more accurate than a method using AdaBoost.