{CSI}{NN}: Reverse engineering of neural network architectures through electromagnetic side channel
L Batina, S Bhasin, D Jap, S Picek - 28th USENIX Security Symposium …, 2019 - usenix.org
Machine learning has become mainstream across industries. Numerous examples prove the
validity of it for security applications. In this work, we investigate how to reverse engineer a …
validity of it for security applications. In this work, we investigate how to reverse engineer a …
CSI neural network: Using side-channels to recover your artificial neural network information
L Batina, S Bhasin, D Jap, S Picek - arXiv preprint arXiv:1810.09076, 2018 - arxiv.org
Machine learning has become mainstream across industries. Numerous examples proved
the validity of it for security applications. In this work, we investigate how to reverse engineer …
the validity of it for security applications. In this work, we investigate how to reverse engineer …
On reverse engineering neural network implementation on GPU
Ł Chmielewski, L Weissbart - … , AIBlock, AIHWS, AIoTS, CIMSS, Cloud S&P …, 2021 - Springer
In recent years machine learning has become increasingly mainstream across industries.
Additionally, Graphical Processing Unit (GPU) accelerators are widely deployed in various …
Additionally, Graphical Processing Unit (GPU) accelerators are widely deployed in various …
Simple electromagnetic analysis against activation functions of deep neural networks
G Takatoi, T Sugawara, K Sakiyama, Y Li - Applied Cryptography and …, 2020 - Springer
From cloud computing to edge computing, the deployment of artificial intelligence (AI) has
been evolving to fit a wide range of applications. However, the security over edge AI is not …
been evolving to fit a wide range of applications. However, the security over edge AI is not …
Incremental inference for probabilistic programs
M Cusumano-Towner, B Bichsel, T Gehr… - Proceedings of the 39th …, 2018 - dl.acm.org
We present a novel approach for approximate sampling in probabilistic programs based on
incremental inference. The key idea is to adapt the samples for a program P into samples for …
incremental inference. The key idea is to adapt the samples for a program P into samples for …
A targeted privacy-preserving data publishing method based on Bayesian network
Z Zhou, Y Wang, X Yu, J Miao - IEEE Access, 2022 - ieeexplore.ieee.org
Privacy-preserving data publishing (PPDP) is an essential prerequisite for data-driven AI
technologies,(such as data mining, machine learning, deep learning, etc.) to extract …
technologies,(such as data mining, machine learning, deep learning, etc.) to extract …
Exact and Efficient Bayesian Inference for Privacy Risk Quantification
RC Rønneberg, R Pardo, A Wąsowski - International Conference on …, 2023 - Springer
Data analysis has high value both for commercial and research purposes. However,
disclosing analysis results may pose severe privacy risk to individuals. Privug is a method to …
disclosing analysis results may pose severe privacy risk to individuals. Privug is a method to …
The limits of SEMA on distinguishing similar activation functions of embedded deep neural networks
G Takatoi, T Sugawara, K Sakiyama, Y Hara-Azumi… - Applied Sciences, 2022 - mdpi.com
Artificial intelligence (AI) is progressing rapidly, and in this trend, edge AI has been
researched intensively. However, much less work has been performed around the security …
researched intensively. However, much less work has been performed around the security …
Privug: using probabilistic programming for quantifying leakage in privacy risk analysis
R Pardo, W Rafnsson, CW Probst… - European Symposium on …, 2021 - Springer
Disclosure of data analytics results has important scientific and commercial justifications.
However, no data shall be disclosed without a diligent investigation of risks for privacy of …
However, no data shall be disclosed without a diligent investigation of risks for privacy of …
Efficient Synthesis with Probabilistic Constraints
S Drews, A Albarghouthi, L D'Antoni - … , CAV 2019, New York City, NY, USA …, 2019 - Springer
We consider the problem of synthesizing a program given a probabilistic specification of its
desired behavior. Specifically, we study the recent paradigm of distribution-guided inductive …
desired behavior. Specifically, we study the recent paradigm of distribution-guided inductive …