Secure and Efficient Protocols for Iris and Fingerprint Identification

摘要: Recent advances in biometric recognition and the increasing use of data prompt significant privacy challenges associated with possible misuse, loss, or theft data. Biometric matching is often performed by two mutually distrustful parties, one which holds image while other owns a possibly large collection. Due to liability considerations, neither party willing share its This gives rise need develop secure computation techniques over where no information revealed parties except outcome comparison search. To address problem, this work we implement first privacy-preserving identification protocol for iris codes. We also design fingerprint based on FingerCodes substantial improvement performance compared existing solutions. show that new optimizations employed allow us achieve particularly efficient protocols suitable sets obtain notable gain state-of-the-art prior work.