On 9/11/13 12:08 PM, Eugen Leitl wrote: > On Wed, Sep 11, 2013 at 11:04:44AM -0500, Matt Mackall wrote: > >> Similarly, any other sort of one-way algorithm that prevents you from >> reconstructing a valid input from the stored data is not going to work. > > Typical fingerprint matching uses classification, recognizing and > encoding multiple features into a vector. You could use a one-way > hash on that vector. This is likely subject to a precompiled hash > lookup table attack, as the number of all possible fingerprints, > quantized via a classification vector is not that large.
There's a good deal of existing research out there on using symmeteric hashes -- a hash that can accept discrete inputs in arbitrary order and always calculate to the same value -- for secure biometric template storage and matching. Here is a paper I point people to that many of you will find absolutely fascinating (although it's been some years so do check citations pointing to this for further work): Sergey Tulyakov, Faisal Farooq, Praveer Mansukhani, & Venu Govindaraju. (2007). Symmetric hash functions for secure fingerprint biometric systems. Pattern Recognition Letters, 28(16), 2427–2436. Retrieved from http://www.researchgate.net/publication/222570842_Symmetric_hash_functions_for_secure_fingerprint_biometric_systems/file/79e4150d06419e02ec.pdf -- Joseph Lorenzo Hall Senior Staff Technologist Center for Democracy & Technology 1634 I ST NW STE 1100 Washington DC 20006-4011 (p) 202-407-8825 (f) 202-637-0968 j...@cdt.org PGP: https://josephhall.org/gpg-key fingerprint: BE7E A889 7742 8773 301B 4FA1 C0E2 6D90 F257 77F8 -- Liberationtech is a public list whose archives are searchable on Google. Violations of list guidelines will get you moderated: https://mailman.stanford.edu/mailman/listinfo/liberationtech. Unsubscribe, change to digest, or change password by emailing moderator at compa...@stanford.edu.
