Transfer Learning for Iris Recognition with Non-Linear Scaling

Abstract

In this work, we intend to investigate the application of transferred VGG19-based CNN model to solve the problem of iris recognition. By employing non-linear scaling technique, the transferred VGG19-based CNN model is modified, re-trained, and finely tuned for iris recognition. The performance of each architecture is assessed by an experiment conducted onIIT Delhi database. Reported results indicated that the best recognition rate can be achieved by employing non-linear scaling operation in VGG19-based CNN model.