Neurogenetic reconstruction of biometric templates: A new security threat?

摘要

In this paper, we demonstrate how neurogenetic reconstruction can be used to reconstruct facial images from biometric templates extracted using Local Binary Patterns (LBPs). We also demonstrate the process of neurogenetic distortion of biometric templates in order to mitigate neurogenetic reconstruction. Our results show that reconstructed images can be used to recognize individuals within a dataset with a high degree of accuracy. Our results also show that neurogenetic distortion can be used to successfully distort biometric templates to prevent image reconstruction.