Score Level Fusion for Iris and Periocular Biometrics Recogniton Based on Deep Learning Cover

.blurhash-client-img { display: none !important; }

Score Level Fusion for Iris and Periocular Biometrics Recogniton Based on Deep Learning

International Journal of Advanced Network, Monitoring and Controls

Volume 7 (2022): Issue 4 (January 2022)

By: Yufei Wang, Songze Lei, Yonggang Li, Bo Liu and Huan Zuo

Open Access

|May 2023

Figures & Tables

Iris pre-processing diagram

Iris normalisation

Schematic diagram of scorer layer fusion

Fusion plot of different weights on CASIA-Iris-Lamp

Schematic representation of the fusion of different weights on CASIA-Iris-Distance

Comparison results of different weights

DataSet	CASIA-Iris-Lamp	CASIA-Iris-Distance
1:9	0.968541	0.935883
2:8	0.971083	0.935883
3:7	0.976167	0.939257
4:6	0.982840	0.9392575
5:5	0.992373	0.9617547
6:4	0.990149	0.917885
7:3	0.986653	0.907761
8:2	0.983158	0.897637
9:1	0.980934	0.894263

Experimental configuration

Relevant parameters	Parameter settings
input_shape	224×224×3
batch_size	32
epoch	2000
optimizers	rmsprop
learning_rate	10⁻³
rho	0.9
decay	10-6

Environment configuration

Hardware Environment	ProcessorGraphics card	Inter(R) XeonGold b6254Navidia GeForce RTX2080Ti
Software Environment	Operating SystemFrameworkProgramming languageCompiler	Linux Ubuntu 16.04Tensorflow-GPUPython3.6.8PyCharm

Comparison of experimental results

Dataset	Fusion rule	Recognition accuracy
CASIA-Iris-Lamp	Maximum value	0.992055
	Minimum value	0.955830
	Multiplication	0.992017
	Additive	0.992373
	Unimodal periocular	0.982617
	Unimodal iris	0.981252
	Adaptive weighting	0.993831

CASIA-Iris-Distance	Maximum value	0.961754
	Minimum value	0.866141
	Multiplication	0.972512
	Additive	0.961754
	Unimodal periocular	0.933633
	Unimodal iris	0.894263
	Adaptive weighting	0.985376

Experimental data sets

Dataset	Category	Iris images	Periocular images	Training set	Validation set	Test set
CASIA-Iris-Lamp	791	15820	15820	9492	3164	3164
CASIA-Iris-Distance	284	5073	5073	3160	1024	889

DOI: https://doi.org/10.2478/ijanmc-2022-0033 | Journal eISSN: 2470-8038

Journal RSS Feed

Language: English

Page range: 21 - 30

Published on: May 26, 2023

Published by: Xi’an Technological University

In partnership with: Paradigm Publishing Services

Publication frequency: 4 issues per year

Keywords:

Multimodal Fusion,

Iris Recognition,

Periocular Recognition,

Score Level Fusion

Related subjects:

Computer sciences,

Computer sciences, other

© 2023 Yufei Wang, Songze Lei, Yonggang Li, Bo Liu, Huan Zuo, published by Xi’an Technological University
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Previous article Volume 7 (2022): Issue 4 (January 2022)Next article