An Algorithmic Approach for Optimising Biometric Systems Using Liveness and Coercion Detection

This paper looks at the possibility of creating an algorithm that will combine liveness and coercion modalities, along with organisational factors such as workforce composition. The hypothesis is that the algorithm can produce a value that can further be used to compare different setups of biometric security for self-optimisation by taking in the context of technique compatibility and user requirements. To this end, the algorithm focuses on four main aspects: time, participants, anomalous user-bases, and device redundancy inside a typical organisation. An experimental methodology has been used, focusing on the development of the algorithm, its associated effects, and how different parameters can be reliably estimated. After testing, the algorithm is proved to work as it creates an appropriate value, called the security value, which can be used to discover the best combinations of modalities for fusion development or practical installation for a given situation. There are some issues with this primarily due to data provision, the requirements for more data to parse through the algorithm, and finally, the need for a suitable interface, otherwise it may be too complex for efficient usage in a traditional security environment. There are potential implications within a general security application such as liveness and coercion multimodal fusion and autonomous system development and pervasive environments, allowing dynamic security systems to be developed. However, the main focus of this algorithm is to highlight the fusion of liveness and coercion detection and how they can be best applied to specific security scenarios.