A Capsule Network-Based Method for Detecting Finger Vein Impersonation Attacks

By improving capsule networks and Bayesian routing algorithms, the accuracy and adaptability of finger vein spoofing attack detection are enhanced, solving the problems of accuracy and rotation sensitivity under small sample datasets, and achieving efficient identification of genuine and fake veins.

CN116246355BActive Publication Date: 2026-01-30NANJING UNIV OF POSTS & TELECOMM
1 Cites 0 Cited by

Patent Information

Application Number
CN202310056814.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-01-18
Publication Date
2026-01-30
Estimated Expiration
2043-01-18

AI Technical Summary

Technical Problem

Existing finger vein spoofing detection methods are not very accurate on small sample datasets, have poor sensitivity to finger displacement and rotation, require a large number of training samples for neural networks, and cannot effectively distinguish between real and fake vein images.

Method used

An improved capsule network is adopted, combined with a Bayesian routing algorithm, to measure the data point concentration by voting consistency among capsules and variational posterior differential entropy, thereby improving classification accuracy. It is suitable for finger vein spoofing attack detection on small sample datasets.

Benefits of technology

It improves the accuracy of classifying true and false veins, reduces training errors, enhances the network's adaptability to finger offset and rotation scenarios, and reduces dependence on additional devices and computing resources.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

This paper proposes a method for detecting finger vein spoofing attacks based on capsule networks. Capsule networks are not only suitable for small-sample finger vein datasets, but also, by replacing neurons with vector-represented capsules on top of CNNs, they can better handle spatial information such as relative position and angle, enhancing the network's adaptability to finger offset and rotation scenarios. A Bayesian routing algorithm is proposed, incorporating the differential entropy of the capsules as a consideration in calculating activation values. During final classification, feature capsules with high activation probabilities and high concentration are selected, which helps improve the accuracy of classifying genuine and fake veins. By simulating the uncertainty of capsule parameters, training errors can be reduced, improving recognition accuracy.
Need to check novelty before this filing date? Find Prior Art