English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

With deep learning deployed in many security-sensitive areas, machinelearning security is becoming progressively important. Recent studiesdemonstrate attackers can exploit system-level techniques exploiting theRowHammer vulnerability of DRAM to deterministically and precisely flip bits inDeep Neural Networks (DNN) model weights to affect inference accuracy. Theexisting defense mechanisms are software-based, such as weight reconstructionrequiring expensive training overhead or performance degradation. On the otherhand, generic hardware-based victim-/aggressor-focused mechanisms imposeexpensive hardware overheads and preserve the spatial connection between victimand aggressor rows. In this paper, we present the first DRAM-basedvictim-focused defense mechanism tailored for quantized DNNs, namedDNN-Defender that leverages the potential of in-DRAM swapping to withstand thetargeted bit-flip attacks with a priority protection mechanism. Our resultsindicate that DNN-Defender can deliver a high level of protection downgradingthe performance of targeted RowHammer attacks to a random attack level. Inaddition, the proposed defense has no accuracy drop on CIFAR-10 and ImageNetdatasets without requiring any software training or incurring hardwareoverhead.

DNN-Defender: A Victim-Focused In-DRAM Defense Mechanism for Taming Adversarial Weight Attack on DNNs