The finger-vein-based deep neural network authentication system has been applied widely in real scenarios, such as countries' banking and entrance guard systems. However, to ensure performance, the deep neural network should train many parameters, which needs lots of time and computing resources. This paper proposes a method that introduces artificial features with prior knowledge into the convolution layer. First, it designs a multi-direction pattern base on the traditional local binary pattern, which extracts general spatial information and also reduces the spatial dimension. Then, establishes a sample effective deep convolutional neural network via combination with convolution, with the ability to extract deeper finger vein features. Finally, trains the model with a composite loss function to increase the inter-class distance and reduce the intra-class distance. Experiments show that the proposed methods achieve a good performance of higher stability and accuracy of finger vein recognition.
Huijie ZHANG
Southeast University
Ling LU
Southeast University,Nanjing Medical University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Huijie ZHANG, Ling LU, "Local Binary Convolution Based Prior Knowledge of Multi-Direction Features for Finger Vein Verification" in IEICE TRANSACTIONS on Information,
vol. E106-D, no. 5, pp. 1089-1093, May 2023, doi: 10.1587/transinf.2022EDL8095.
Abstract: The finger-vein-based deep neural network authentication system has been applied widely in real scenarios, such as countries' banking and entrance guard systems. However, to ensure performance, the deep neural network should train many parameters, which needs lots of time and computing resources. This paper proposes a method that introduces artificial features with prior knowledge into the convolution layer. First, it designs a multi-direction pattern base on the traditional local binary pattern, which extracts general spatial information and also reduces the spatial dimension. Then, establishes a sample effective deep convolutional neural network via combination with convolution, with the ability to extract deeper finger vein features. Finally, trains the model with a composite loss function to increase the inter-class distance and reduce the intra-class distance. Experiments show that the proposed methods achieve a good performance of higher stability and accuracy of finger vein recognition.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2022EDL8095/_p
Copy
@ARTICLE{e106-d_5_1089,
author={Huijie ZHANG, Ling LU, },
journal={IEICE TRANSACTIONS on Information},
title={Local Binary Convolution Based Prior Knowledge of Multi-Direction Features for Finger Vein Verification},
year={2023},
volume={E106-D},
number={5},
pages={1089-1093},
abstract={The finger-vein-based deep neural network authentication system has been applied widely in real scenarios, such as countries' banking and entrance guard systems. However, to ensure performance, the deep neural network should train many parameters, which needs lots of time and computing resources. This paper proposes a method that introduces artificial features with prior knowledge into the convolution layer. First, it designs a multi-direction pattern base on the traditional local binary pattern, which extracts general spatial information and also reduces the spatial dimension. Then, establishes a sample effective deep convolutional neural network via combination with convolution, with the ability to extract deeper finger vein features. Finally, trains the model with a composite loss function to increase the inter-class distance and reduce the intra-class distance. Experiments show that the proposed methods achieve a good performance of higher stability and accuracy of finger vein recognition.},
keywords={},
doi={10.1587/transinf.2022EDL8095},
ISSN={1745-1361},
month={May},}
Copy
TY - JOUR
TI - Local Binary Convolution Based Prior Knowledge of Multi-Direction Features for Finger Vein Verification
T2 - IEICE TRANSACTIONS on Information
SP - 1089
EP - 1093
AU - Huijie ZHANG
AU - Ling LU
PY - 2023
DO - 10.1587/transinf.2022EDL8095
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E106-D
IS - 5
JA - IEICE TRANSACTIONS on Information
Y1 - May 2023
AB - The finger-vein-based deep neural network authentication system has been applied widely in real scenarios, such as countries' banking and entrance guard systems. However, to ensure performance, the deep neural network should train many parameters, which needs lots of time and computing resources. This paper proposes a method that introduces artificial features with prior knowledge into the convolution layer. First, it designs a multi-direction pattern base on the traditional local binary pattern, which extracts general spatial information and also reduces the spatial dimension. Then, establishes a sample effective deep convolutional neural network via combination with convolution, with the ability to extract deeper finger vein features. Finally, trains the model with a composite loss function to increase the inter-class distance and reduce the intra-class distance. Experiments show that the proposed methods achieve a good performance of higher stability and accuracy of finger vein recognition.
ER -