all InfoSec news
Diffusion Models For Stronger Face Morphing Attacks. (arXiv:2301.04218v1 [cs.CV])
cs.CR updates on arXiv.org arxiv.org
Face morphing attacks seek to deceive a Face Recognition (FR) system by
presenting a morphed image consisting of the biometric qualities from two
different identities with the aim of triggering a false acceptance with one of
the two identities, thereby presenting a significant threat to biometric
systems. The success of a morphing attack is dependent on the ability of the
morphed image to represent the biometric characteristics of both identities
that were used to create the image. We present a …
aim attack attacks biometric diffusion models face recognition identities recognition system systems threat