Dr Abdarahmane WONE

Dr. Wone obtained his Ph.D. from University of Caen-Normandie in 2023 under the supervision of Pr Christophe Rosenberger and Pr. Christophe Charrier.
He carried out his researches at the GREYC LAB in the SAFE (Security Architecture Forensic biomEtrics) team. He is an associate researcher at the GREYC lab/SAFE team

CV

English

French

Publications

International conferences

  • A. Wone, J. Di Manno, C. Rosenberger and C. Charrier, ”Capture biases in Fingerprint Systems,” 2023 International Conference on Cyberworlds (CW), 2023
  • A. Wone, J. Di Manno, C. Rosenberger and C. Charrier, ”Digitally Synthetized Fingerprint Spoofs: A Threat For Anti-Spoofing Systems?,” 2022 International Conference on Cyberworlds (CW), 2022, pp. 255-258
  • A. Wone, J. D. Manno, C. Charrier and C. Rosenberger, ”Impact Of Environmental Conditions On Fingerprint Systems Performance,” 2021 18th International Conference on Privacy, Security and Trust (PST), 2021, pp. 1-5

Thesis

Contribution to the certification of fingerprint systems: Towards the reproducibility of the evaluation:
Computer science is more and more present in our daily lives for multiple tasks. Considering its multiple services and applications, its security is an essential guarantee of the trust that a final user can have. Biometrics is a solution for that. Biometrics gathers all the methods and techniques to automatically recognize individuals thanks to their physical, biological, or behavioral traits. In authentication or identification tasks, it allows to be sure that only legitimate people have access to some data, places, or applications, etc. Whether it is for daily tasks such as unlocking our phones or some specific tasks as validating a banking operation, biometrics is more and more used not only to secure some solutions but also to simplify our lives.

However, we must be sure about the security of biometric systems. A biometric system should be able to recognize a genuine user, reject an illegitimate and resist to attacks. My thesis deals with the certification of biometric fingerprint systems which is an important step prior to the production and deployment of a biometric solution. We explore the factors impacting the behavior of a fingerprint system in its recognition capacities. Indeed, a test of a biometric system should not be dependent on the conditions it was tested. We want biometric testing to be repeatable and biometric solutions to be interoperable. We analyze how environmental conditions as well as the acquisition technology can impact the recognition of a biometric fingerprint system. Moreover, the testing of resistance to attack has legal and operational constraints that make the testing difficult. We propose to explore the generation of synthetic attack instruments using a deep-leaning solution for the evaluation of biometric fingerprint solutions. We validate our studies with an objective method that we propose in the dissertation.

The manuscript of my thesis can be downloaded here.

Vulgalisation

  • Joël Di Manno, Jean Fang et Abdarahmane Wone, Using synthetic data to fight fingerprint spoofing, Biometric Technology Today, Vol. 2023, No. 4,ISSN 1873-1880
  • A. WONE, J. DIMANNO, C. ROSENBERGER, C. CHARRIER, impact ofenvironmental conditions on fingerprint systems, Poster, journée dulaboratoire GREYC, IMEC, 2022.
  • A. WONE, J. DIMANNO, C. ROSENBERGER, C. CHARRIER, Artificial Intelligence inthe certification of biometric systems, Poster, Inria-DFKI European SummerSchool on AI, saarbrücken, 2022

Blog articles

Research

My research themes are:

  • Deep learning
  • Artificial Intelligence
  • Biometrics
  • Quality of biometric data
  • Synthetic data