Neural Fuzzy Extractors: A Secure Way to Use Artificial Neural Networks for Biometric User Authentication

Powered by new advances in sensor development and artificial intelligence, the decreasing cost of computation, pervasiveness handheld computation devices, biometric user authentication (and identification) is rapidly becoming ubiquitous. Modern approaches to authentication, based on sophisticated machine learning techniques, cannot avoid storing either trained-classifier details or explicit data, thus exposing users’ credentials falsification. In this paper, we introduce a secure way handle user-specific information involved with use neural networks for authentication. Our proposed architecture, called Neural Fuzzy Extractor (NFE), allows coupling pre-existing classifiers fuzzy extractors, through an artificial-neuralnetwork-based buffer expander, minimal no performance degradation. The NFE offers all advantages modern deep-learningbased security standard extractors. We demonstrate retrofit few classic networks, simple scenarios.