Fuzzy Commitments Offer Insufficient Protection to Biometric Templates Produced by Deep Learning (Record no. 644)

000 -LEADER
fixed length control field 01425nam a2200169 a 4500
005 - DATE AND TIME OF LATEST TRANSACTION
control field 20260901030330.0
008 - FIXED-LENGTH DATA ELEMENTS--GENERAL INFORMATION
fixed length control field 250101s2020 xx o 000 0 eng d
100 1# - MAIN ENTRY--PERSONAL NAME
Personal name Danny Keller
245 10 - TITLE STATEMENT
Title Fuzzy Commitments Offer Insufficient Protection to Biometric Templates Produced by Deep Learning
264 #1 - PRODUCTION, PUBLICATION, DISTRIBUTION, MANUFACTURE, AND COPYRIGHT NOTICE
Name of producer, publisher, distributor, manufacturer arXiv
Date of production, publication, distribution, manufacture, or copyright notice 2020
336 ## - CONTENT TYPE
Content type term text
338 ## - CARRIER TYPE
Carrier type term online resource
520 ## - SUMMARY, ETC.
Summary, etc. In this work, we study the protection that fuzzy commitments offer when they are applied to facial images, processed by the state of the art deep learning facial recognition systems. We show that while these systems are capable of producing great accuracy, they produce templates of too little entropy. As a result, we present a reconstruction attack that takes a protected template, and reconstructs a facial image. The reconstructed facial images greatly resemble the original ones. In the simplest attack scenario, more than 78% of these reconstructed templates succeed in unlocking an account (when the system is configured to 0.1% FAR). Even in the "hardest" settings (in which we take a reconstructed image from one system and use it in a different system, with different feature extraction process) the reconstructed image offers 50 to 120 times higher success rates than the system's FAR.
506 0# - RESTRICTIONS ON ACCESS NOTE
Terms governing access Open access — freely available to read.
856 40 - ELECTRONIC LOCATION AND ACCESS
Uniform Resource Identifier <a href="https://arxiv.org/pdf/2012.13293v1">https://arxiv.org/pdf/2012.13293v1</a>
Link text Read the full paper (PDF)
942 ## - ADDED ENTRY ELEMENTS (KOHA)
Koha item type Electronic resource (link)
Holdings
Withdrawn status Lost status Damaged status Not for loan Collection Home library Current library Shelving location Date acquired Total checkouts Barcode Date last seen Uniform resource identifier Price effective from Koha item type
      Available online Cybersecurity Yegates University Library Yegates University Library Science and Computing 09/01/2026   YGE000905 09/01/2026 https://arxiv.org/pdf/2012.13293v1 09/01/2026 Research paper — read online