DeepNcode: Encoding-Based Protection against Bit-Flip Attacks on Neural Networks

arXiv:2405.13891v1 Announce Type: new
Abstract: Fault injection attacks are a potent threat against embedded implementations of neural network models. Several attack vectors have been proposed, such as misclassification, model extraction, and trojan/backdoor planting. Most of these attacks work by flipping bits in the memory where quantized model parameters are stored.
In this paper, we introduce an encoding-based protection method against bit-flip attacks on neural networks, titled DeepNcode. We experimentally evaluate our proposal with several publicly available models and datasets, by …

arxiv attack attacks attack vectors backdoor bits cs.ai cs.cr embedded encoding extraction flip injection injection attacks memory model extraction network networks neural network neural networks protection threat trojan work