Transient liveness factor an attempt towards identification of innate fingerprint under varying spoof sample
Akhilesh Verma, Vijay Kumar Gupta, Savita Goel · International Journal of Advanced Intelligence Paradigms · 2023
Fingerprint liveness detection works reasonably well when attacks are known. A 'constrained' design is biased to the taste of the researcher. Solutions appreciated initially, discarded later on based on issues of reproducibility and interoperability. Still, the scope of abstracting the solution in a wider viewpoint is missing. This work underlines the inherent limitations of current practices, specifically, those encountered under 'seen' conditions, where limited training data is available. We envision this work towards the understanding model of innateness 'a non-separable reality' of live samples instead of learning non-exhaustive spoof samples. We highlight concepts and efforts for a near-ideal spoof detection system. Future applications will heavily rely on trustworthy biometric-system, so an urgent need to interpret and standardise the fingerprint presentation attack detection (FPAD) system is required. A proposal is put forward to model the liveness using transient liveness factor for future applications based on the fingerprint biometric.