all InfoSec news
Facial Misrecognition Systems: Simple Weight Manipulations Force DNNs to Err Only on Specific Persons. (arXiv:2301.03118v1 [cs.CR])
cs.CR updates on arXiv.org arxiv.org
In this paper we describe how to plant novel types of backdoors in any facial
recognition model based on the popular architecture of deep Siamese neural
networks, by mathematically changing a small fraction of its weights (i.e.,
without using any additional training or optimization). These backdoors force
the system to err only on specific persons which are preselected by the
attacker. For example, we show how such a backdoored system can take any two
images of a particular person and …
architecture backdoors changing facial facial recognition networks neural networks novel optimization popular recognition simple system systems training types