Keystroke Dynamics With Low Constraints SVM Based Passphrase Enrollment

Romain Giot (1) , Mohamad El-Abed (1) , Christophe Rosenberger (1)

1 Equipe SAFE - Laboratoire GREYC - UMR6072

Keystroke dynamics biometric systems have been studied for more than twenty years. They are very well perceived by users, they may be one of the cheapest biometric system (as no specific material is required) even if they are not commonly spread and used. We propose in this paper a new method based on SVM learning satisfying operational conditions (no more than 5 captures for the enrollment step). In the proposed method, users are authenticated thanks to keystroke dynamics of a passphrase (that can be chosen by the system administrator). We use the GREYC keystroke benchmark that is composed of a large number of users (100) for validation purposes. We tested the proposed method face to four other methods from the state of the art. Experimental results show that the proposed method outperforms them in an operational context.

Mots clés: biométrie; dynamique de frappe au clavier; svm

Domaines: Apprentissage [cs.LG]; Autre [cs.OH]

Romain Giot, Mohamad El-Abed, Christophe Rosenberger. Keystroke Dynamics With Low Constraints SVM Based Passphrase Enrollment. IEEE International Conference on Biometrics: Theory, Applications and Systems (BTAS 2009), Sep 2009, Washington, United States. pp.6, ⟨10.1109/BTAS.2009.5339028⟩. ⟨hal-00432775⟩