Complex-Valued Iris Recognition Network.

Kien Nguyen,Clinton Fookes,Sridha Sridharan,Arun Ross

doi:10.1109/tpami.2022.3152857

Abstract

In this work, we design a fully complex-valued neural network for the task of iris recognition. Unlike the problem of general object recognition, where real-valued neural networks can be used to extract pertinent features, iris recognition depends on the extraction of both phase and magnitude information from the input iris texture in order to better represent its biometric content. This necessitates the extraction and processing of phase information that cannot be effectively handled by a real-valued neural network. In this regard, we design a fully complex-valued neural network that can better capture the multi-scale, multi-resolution, and multi-orientation phase and amplitude features of the iris texture. We show a strong correspondence of the proposed complex-valued iris recognition network with Gabor wavelets that are used to generate the classical IrisCode; however, the proposed method enables a new capability of automatic complex-valued feature learning that is tailored for iris recognition. We conduct experiments on three benchmark datasets - ND-CrossSensor-2013, CASIA-Iris-Thousand and UBIRIS.v2 - and show the benefit of the proposed network for the task of iris recognition. We exploit visualization schemes to convey how the complex-valued network, when compared to standard real-valued networks, extracts fundamentally different features from the iris texture.

Highlights

The last few years have seen a transition in the iris recognition community to deep neural networks to take advantage of their automatic feature learning capability [24], [31], [52], [55]
The fact that the iris texture is stochastic [11] with no consistent shapes, edges or semantic structures would make real-valued networks struggle to learn any meaningful shapes from the iris texture and not to realize the full potential of automatic feature learning in the iris recognition setting
The fact that the iris texture in iris recognition has no consistent shapes, edges, or semantic structure unlike in classical object detection and classification, causes realvalued networks to struggle to learn any meaningful shapes from the iris texture and is unable to realize the full potential of automatic feature learning in the iris recognition setting

Summary

Introduction

T HE human iris is a powerful biometric pattern that has the potential to deliver high recognition accuracy at low false match rates. The last few years have seen a transition in the iris recognition community to deep neural networks to take advantage of their automatic feature learning capability [24], [31], [52], [55]. This provides us with an alternative approach in feature design by automatically learning and discovering feature representations directly from data, eliminating some of the pitfalls in developing handcrafted features [2]. Despite the promise and a number of initial efforts in this direction, deep neural networks have not been exactly revolutionary in iris recognition This could be because existing deep iris recognition networks in the literature are directly derived from general deep learning theory for natural images.

Methods

Results

Conclusion