Poster: gait-based smartphone user identification

Biometric signatures based on a user’s physical features form a natural approach to authentication, as they are often high in entropy and hard to forge or forget. Gait, the rhythmical body movements of human beings as they walk, is a particularly appealing biometric because of its distinguishability [2]. In this work, we provide a proof-of-concept implementation of gait-based user identification using the tri-axial accelerometer sensor of the Motorola Droid phone. It is envisioned that this means of user identification could be utilized as a component of a larger recognition system, one that might fully amalgamate all relevant sensed data to generate an idea of whether a legitimate user has possession of the phone at any given instant. 1