Hardware side-channel attacks can recover deep-learning model architecture, parameters and inputs; this survey taxonomizes the leaks, attacks and countermeasures.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Securing Deep Learning Hardware: A Survey of Side-Channel Vulnerabilities and Countermeasures
Hardware side-channel attacks can recover deep-learning model architecture, parameters and inputs; this survey taxonomizes the leaks, attacks and countermeasures.