In a verifiable computation system, a client outsources a computation to a third party while keeping the confidence in the result. In the context of biometrics, when data is outsourced to a cloud provider, cloud users can require the cloud to perform some operations over their data. It is of particular interest to rely on new integrity mechanisms which ensure the correct execution of outsourced computations. A malicious or a lazy cloud provider could indeed modify the process to gain some advantages as, for instance, reducing the cost of operating a system. Thus verifiable computation is a tool to enhance the security of a system when parts of the data are outsourced. It will be very helpful for use cases like a mobile client that needs to rely on a cloud provider for authentication decision.
On the one hand, some ad-hoc techniques could be designed for a specific operation, as for instance in , that relies on the use of carefully chosen and/or controlled data from the database. On the other hand, recently, generic constructions that become nearly practical have been suggested in the cryptographic community, such as Pepper , Pinocchio, Geppeto [3, 4], or Pantry , to name a few. Interestingly, a part of these works also enable some kind of confidentiality constraints. For biometric systems, this may allow to deal with secrecy of part of the data or parameters.
The global objective of the PhD proposal is to assess the feasibility of the existing verifiable computation systems, or by adapting them into new solutions, in the context of biometric verification for different use cases.
 M. Blanton, Y. Zhang, and K. B. Frikken. Secure and verifiable outsourcing of large-scale biometric computations, ACM Trans. Inf. Syst. Secur., vol. 16, no. 3, p. 11, 2013.
 S. T. V. Setty, R. McPherson, A. J. Blumberg, and M. Walfish. Making argument systems for outsourced computation practical (sometimes), in 19th Annual Network and Distributed System Security Symposium, NDSS, 2012.
 B. Parno, J. Howell, C. Gentry, and M. Raykova. Pinocchio: Nearly practical verifiable computation, in IEEE Symposium on Security and Privacy, 2013, pp. 238_252.
 Craig Costello, Cedric Fournet, Jon Howell, Markulf Kohlweiss, Benjamin Kreuter, Michael Naehrig, Bryan Parno, and Samee Zahur. Gepetto: Versatile Verifiable Computation, in S&P2015 (36th IEEE Symposium on Security and Privacy, San Jose, California, May 18-20, 2015)
 B. Braun, A. J. Feldman, Z. Ren, S. T. V. Setty, A. J. Blumberg, and M. Walfish. Verifying computations with state, in ACM SIGOPS 24th Symposium on Operating Systems Principles, SOSP, 2013, pp. 341_357.
This PhD will be supervised by two partners under the auspices of the european project TREDISEC starting in 2015: Eurecom for the cryptographic techniques and Morpho for the application to biometric matching and the integration of the findings in an industrial environment.
The thesis will be located at Morpho’s premises (Issy-Les-Moulineaux, France), with frequent interactions with the other partner (Eurecom, Sophia-Antipolis, France). Participation of the candidate to the life of the project TREDISEC is also required.
The candidate should have a Master’s degree in computer science or a related field. Knowledge and motivation for one of the following fields would be appreciated: security, cryptography, software development.
Knowledge of French is not required.
Supervisor: Refik Molva (Eurecom), Refik.Molva@eurecom.fr
Co-supervisor: Hervé Chabanne (Morpho), firstname.lastname@example.org