Dual-channel heterogeneous network for finger vein recognition based on transfer learning and improved coordinate attention
by Yuchuan Chen; Jing Jie; Jiahui Chai; Hui Zheng; Xiaoli Wu
International Journal of Computing Science and Mathematics (IJCSM), Vol. 19, No. 4, 2024

Abstract: In order to effectively improve the recognition accuracy of finger veins, a dual-channel heterogeneous network based on transfer learning and coordinate attention is proposed. Firstly, a VGG16 and a DenseNet121 pretrained by ImageNet are selected to form the dual-channel heterogeneous network, and then the new network is trained again by finger vein image set to validly improve its characterisation power without too much training cost. Secondly, an improved coordinate attention is introduced to help the network evaluate the importance of information in different spatial locations, which contributes to outputting valid finger vein features to the classification layer. Finally, the network performance was tested based on a 700-type of finger vein image set. The experimental results show that transfer learning and improved coordinate attention validly enhance the recognition effect with low computation complexity, and the proposed network shows excellent performance for finger vein recognition.

Online publication date: Wed, 12-Jun-2024

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Computing Science and Mathematics (IJCSM):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com