CCRNet: a novel data-driven approach to improve cross-domain Iris recognition
- PDF / 2,745,428 Bytes
- 25 Pages / 439.37 x 666.142 pts Page_size
- 51 Downloads / 166 Views
CCRNet: a novel data-driven approach to improve cross-domain Iris recognition Meenakshi Choudhary 1 & Vivek Tiwari 1 & U Venkanna 1 Received: 12 August 2019 / Revised: 11 April 2020 / Accepted: 29 June 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
In spite of the prominence and robustness of iris recognition systems, iris images acquisition using heterogeneous cameras/sensors, is the prime concern in deploying them for wide-scale applications. The textural qualities of iris samples (images) captured through distinct sensors substantially differ due to the differences in illumination and the underlying hardware that yields intra-class variation within the iris dataset. This paper examines three miscellaneous configurations of convolution and residual blocks to improve cross-domain iris recognition. Further, the finest architecture amongst three is identified by the Friedman test, where the statistical differences in proposed architectures are identified based on the outcomes of Nemeny and Bonferroni-Dunn tests. The quantitative performances of these architectures are perceived on several experiments simulated on two iris datasets; ND-CrossSensor-Iris-2013 and ND-iris-0405. The finest model is referred to as “Collaborative Convolutional Residual Network (CCRNet)” and is further examined on several experiments prepared in similar and cross-domains. Results depict that least two error rates reported by CCRNet are 1.06% and 1.21% that enhances the benchmark for the state of the arts. This is due to fast convergence and rapid weights updation achieved from convolution and residual connections, respectively. It helps in recognizing the micro-patterns existing within the iris region and results in better feature discrimination among large numbers of iris subjects. Keywords cross-domain iris recognition . Collaborative Convolutional Residual Network . CCRNet . pre-activation . model selection
* Vivek Tiwari [email protected]
1
Department of Computer Science Engineering, DSPM IIIT Naya Raipur, Naya Raipur, Chhattisgarh 493661, India
Multimedia Tools and Applications
1 Introduction Human identification based on biological features has been extensively used in almost all security control applications. These biological features are characterized by various biometric traits/modalities, such as fingerprints, face, iris, retina, palm-print, DNA, etc. In recent, gait analysis is also increasingly emphasized for person identification. Each of these traits has unique patterns for an individual and is used either alone or in combination with other traits, which is termed as multi-model biometrics [41]. The fingerprints, face and iris recognition are widely used not only in mobile devices but stationary biometric systems as well, e.g. India’s Aadhar program. The facial pattern analysis is performed on 2D and 3D face images [7, 24]. Besides, the iris has also gained vital attention in the past two decades due to its inimitable characteristics such as rich texture with immense information, d
Data Loading...
