As part of a fight against ID misuse, the state of Israel has decided to 
establish a biometric database storing biometric information (fingerprints 
and high definition photographs) of all citizens. Obviously, such a database 
is a privacy and security concern unless properly deployed and handled.

As the prominent publicly acknowledged target of this database is to detect
fraudulent individuals who are in the possession of more than one ID, we
propose two schemes which allows identifying this behavior while offering 
maximal privacy, even in the case of a complete database leakage. 

Our proposed solutions are based on the notion that generating entries
in the database should be controlled, whereas the generated entries should
be as pseudo-random as possible. This way, the generation of the entires
can be done by a practical secure multiparty computation.

This is a joint work with Melissa Chase and with Rita Osadchy.