The Plateau: Imitation Attack Resistance of Gait Biometrics

Constituting a new branch within biometrics, gait biometrics needs to be extensively tested and analyzed to determine its level of fraud resistance. Previous results examining the attack resistance testing of gait authentication systems show that imitation, or mimicking of gait, is a venerable challenge.

This paper presents an experiment where participants are extensively trained to become skilled gait mimickers. Results show that our physiological characteristics tend to work against us when we try to change something as fundamental as the way we walk. Simple gait details can be adopted, but if the imitator changes several characteristics at once, the walk is likely to become uneven and mechanical. The participants showed few indications of learning, and the results of most attackers even worsened over time, showing that training did nothing to help them succeed.

With extensive training an impostor’s performance can change, but this change seems to meet a natural boundary, a limit. This paper introduces the

plateau

, a physiologically predetermined limit to performance, forcing imitators back whenever they attempt to improve further. The location of this plateau determines the outcome of an attack; for success it has to lie below the acceptance threshold corresponding to the Equal Error Rate (EER).