Related Experiment Video
Updated: Aug 7, 2026
08:15
Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
Published on: March 28, 2025
Deep Bio-Hashing Network for Privacy-Preserving Cancelable Finger Vein Recognition
Summary
This study introduces a Deep Bio-Hashing Network (DBHN) for secure finger vein recognition, enhancing privacy by making biometric data cancelable. The method protects against data leakage and misuse, offering robust security for identification systems.
Area of Science:
- Biometrics
- Computer Science
- Information Security
Background:
- Finger vein recognition is crucial for high-security identification.
- Traditional methods risk permanent identity loss from biometric data leakage.
- Cancelable biometrics offer enhanced privacy and reduced misuse risk.
Purpose of the Study:
- To propose a Deep Bio-Hashing Network (DBHN) for end-to-end privacy-preserving and cancelable finger vein recognition.
- To enhance feature alignment and address security issues with tokenized random numbers.
Main Methods:
- Developed a class center alignment module for improved feature alignment.
- Introduced a Deep Bio-Hashing layer using a system-level token for enhanced security.
- Designed a hybrid loss function (classification, localization, class center triplet loss) for DBHN supervision.
Main Results:
- Achieved favorable recognition performance on three public datasets.
- Demonstrated competitive results compared to state-of-the-art hash-based methods.
- Verified cancelable biometrics attributes and resilience against security/privacy attacks.
Conclusions:
- DBHN offers an effective solution for privacy-preserving finger vein recognition.
- The proposed method enhances security by mitigating biometric data leakage risks.
- DBHN provides a robust and resilient approach to cancelable biometrics.