Developing coercion detection solutions for biometrie security
Peter William Matthew, Mark S. Anderson · 2016 SAI Computing Conference (SAI) · 2016
Biometric security depends on its accuracy and efficiency, but is especially vulnerable to spoof attacks. Currently liveness detection has become the standard method of reducing the impact of spoof attacks, however whilst this protects against spoof creation and presentation, it does nothing to detect legitimate users being coerced into accessing systems. This paper looks at this coercion concept, indicating the impact coercion could have on biometric security and how it can be detected. This paper will identify scenarios in which coercion detection could improve security as well as identifying the underlying concepts that make this area up, culminating in the presentation of four high level techniques which can be implemented within a multi-modal biometric system to detect coercion detection.