Abstract:A Two-Factor Authentication (2FA) scheme can authenticate a client if the client is able to provide the possession factor (like biometric feature, smart card) and the knowledge factor (like password, secret key) simultaneously. With only one factor, it is hard for an adversary to impersonate the client to pass the authentication, and thus 2FA provides better security than single-factor authentication schemes. However, as far as we know, all existing 2FA schemes do not consider the leakage of server's database, and their authenticity may fail when the database is also compromised (in addition to one factor). Considering numerous reports of database leakage in the real world, it seems imminent to study and design 2FA schemes resilient to database leakage. In this paper, we formalize security models for 2FA schemes by taking database leakage into account. Our security models consider malicious adversaries who can obtain both the client's one authentication factor and the server's database, and have two requirements, authenticity and zero-knowledge . Authenticity ensures that such malicious adversaries cannot impersonate the client to pass the authentication, while zero-knowledge guarantees that such malicious adversaries obtain no information about the client's the other factor. Zero-knowledge is especially important for biometric features (like faces, fingerprints), which are inherent to human beings and can hardly be changed. Then we propose a biometric-based 2FA scheme with biometric feature and secret key serving as the two authentication factors. Our 2FA scheme has three rounds, and we prove its authenticity and zero-knowledge under database leakage in the random oracle model. Notably, our construction makes a novel use of a recent technical advance called robust property-preserving hashing (Boyle et al., ITCS 2019) together with fully homomorphic encryption, to recognize or discern clients by their biometric samplings in a homomorphic and secure way.

Exploit the Leak: Understanding Risks in Biometric Matchers

Untargeted Near-collision Attacks on Biometrics: Real-world Bounds and Theoretical Limits

General Framework to Evaluate Unlinkability in Biometric Template Protection Systems

On the Risk of Cancelable Biometrics

A False Sense of Privacy: Towards a Reliable Evaluation Methodology for the Anonymization of Biometric Data

Leakage or Identification: Behavior-irrelevant User Identification Leveraging Leakage Current on Laptops

Emerging Biometric Modalities and their Use: Loopholes in the Terminology of the GDPR and Resulting Privacy Risks

Palmsecmatch: A Data-Centric Template Protection Method for Palmprint Recognition

Leakage or Identification

Biometric Systems Private by Design: Reasoning about privacy properties of biometric system architectures

The Multibiometrics Template Protection Algorithm for Lower Quality Iris and Thermal Face Image based on Fuzzy Commitment

Biometric-based two-factor authentication scheme under database leakage

A Survey on Homomorphic Encryption for Biometrics Template Security Based on Machine Learning Models

Theoretical framework for constructing matching algorithms in biometric authentication systems

A secure and private ensemble matcher using multi-vault obfuscated templates

Model-Agnostic Utility-Preserving Biometric Information Anonymization

Fast and Accurate Likelihood Ratio Based Biometric Comparison in the Encrypted Domain

Homomorphic Encryption for Speaker Recognition: Protection of Biometric Templates and Vendor Model Parameters

Systematic Evaluation of Privacy Risks of Machine Learning Models

A novel template protection scheme for multibiometrics based on fuzzy commitment and chaotic system

Privacy-Preserving Biometric System for Secure Fingerprint Authentication