Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AMB-FHE: Adaptive Multi-biometric Fusion with Fully Homomorphic Encryption

Florian Bayer, Christian Rathgeb

TL;DR

Biometric systems face a security-usability trade-off, especially when leveraging multiple modalities. The paper introduces AMB-FHE, an adaptive multi-biometric fusion framework based on the CKKS fully homomorphic encryption scheme that stores concatenated templates in a single ciphertext and performs distance computations entirely in the encrypted domain. The approach supports run-time adaptation via sequential cascaded decision-level fusion, improving usability while enhancing privacy. Experiments on iris and fingerprint data show a fusion EER of about $0.08\%$ and substantial reductions in unnecessary modality presentations, in the range of $72\%$ to $96\%$, highlighting practical impact and guiding future efficiency improvements.

Abstract

Biometric systems strive to balance security and usability. The use of multi-biometric systems combining multiple biometric modalities is usually recommended for high-security applications. However, the presentation of multiple biometric modalities can impair the user-friendliness of the overall system and might not be necessary in all cases. In this work, we present a simple but flexible approach to increase the privacy protection of homomorphically encrypted multi-biometric reference templates while enabling adaptation to security requirements at run-time: An adaptive multi-biometric fusion with fully homomorphic encryption (AMB-FHE). AMB-FHE is benchmarked against a bimodal biometric database consisting of the CASIA iris and MCYT fingerprint datasets using deep neural networks for feature extraction. Our contribution is easy to implement and increases the flexibility of biometric authentication while offering increased privacy protection through joint encryption of templates from multiple modalities.

AMB-FHE: Adaptive Multi-biometric Fusion with Fully Homomorphic Encryption

TL;DR

Biometric systems face a security-usability trade-off, especially when leveraging multiple modalities. The paper introduces AMB-FHE, an adaptive multi-biometric fusion framework based on the CKKS fully homomorphic encryption scheme that stores concatenated templates in a single ciphertext and performs distance computations entirely in the encrypted domain. The approach supports run-time adaptation via sequential cascaded decision-level fusion, improving usability while enhancing privacy. Experiments on iris and fingerprint data show a fusion EER of about and substantial reductions in unnecessary modality presentations, in the range of to , highlighting practical impact and guiding future efficiency improvements.

Abstract

Biometric systems strive to balance security and usability. The use of multi-biometric systems combining multiple biometric modalities is usually recommended for high-security applications. However, the presentation of multiple biometric modalities can impair the user-friendliness of the overall system and might not be necessary in all cases. In this work, we present a simple but flexible approach to increase the privacy protection of homomorphically encrypted multi-biometric reference templates while enabling adaptation to security requirements at run-time: An adaptive multi-biometric fusion with fully homomorphic encryption (AMB-FHE). AMB-FHE is benchmarked against a bimodal biometric database consisting of the CASIA iris and MCYT fingerprint datasets using deep neural networks for feature extraction. Our contribution is easy to implement and increases the flexibility of biometric authentication while offering increased privacy protection through joint encryption of templates from multiple modalities.

Paper Structure

This paper contains 14 sections, 2 equations, 2 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Proposed System
  4. Background
  5. CKKS
  6. Distance Computation
  7. Algorithm
  8. Protocol
  9. Evaluation
  10. Software and Databases
  11. Performance Evaluation
  12. Usability
  13. Security Analysis
  14. Conclusion and Outlook

Figures (2)

  • Figure 1: High-level overview of the AMB-FHE protocol for enrollment and verification with two biometric modalities. Upon enrollment, all supported modalities undergo a series of processes on the client device, including capturing, feature extraction, concatenation, and encryption. During authentication, a previously defined first modality is presented, and further modalities are captured sequentially if and only if the score falls below a certain security threshold.
  • Figure 2: Proposed AMB-FHE protocol for authentication based on Model I (Store on server, compare distributed) concept presented in ISO/IEC 24745.