Manipulating Weights in Face-Recognition AI Systems
Schneier on Security
FEBRUARY 3, 2023
Interesting research: “ Facial Misrecognition Systems: Simple Weight Manipulations Force DNNs to Err Only on Specific Persons “: Abstract: 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).
Let's personalize your content