Hand in Motion: Enhanced authentication through wrist and mouse movement

Behavioral biometrics have been long used as a complementary method to the traditional one-time authentication system. Mouse dynamics, representing an individual's unique patterns of mouse operations, possess a great potential to bridge the security gap between two one-time authentications on the computer. In this paper, we propose a continuous authentication approach by combining the deviceindependent, angle-based mouse movement features and the wrist motion features. Based on a Random Forest Ensemble Classifier (RFEC) and the Sequential Sampling Analysis (SSA), the identity of the user can be continuously verified. Experimental results, based on 26 subjects, show that the proposed approach can reach the False Accept Rate (FAR) of 1.46% and 4.69% for impostors and intruders respectively and a False Reject Rate (FRR) of 0%. Moreover, the proposed approach is proven to be more effective in timely authentication (i.e., making an authentication decision within only 9 to 12 mouse clicks), compared with conventional methods solely based on the mouse geometry and locomotion features.