Privacy preserving layer partitioning for Deep Neural Network models

K Rajasekar, R Loh, KW Fok… - 2024 IEEE Conference …, 2024 - ieeexplore.ieee.org
MLaaS (Machine Learning as a Service) has become popular in the cloud computing
domain, allowing users to leverage cloud resources for running private inference of ML …

Laminator: Verifiable ML Property Cards using Hardware-assisted Attestations

V Duddu, O Järvinen, LJ Gunn, N Asokan - arxiv preprint arxiv …, 2024 - arxiv.org
Regulations increasingly call for various assurances from machine learning (ML) model
providers about their training data, training process, and the behavior of resulting models …

PREFHE, PREFHE-AES and PREFHE-SGX: Secure multiparty computation protocols from fully homomorphic encryption and proxy reencryption with AES and intel …

C Yakupoglu, K Rohloff - … Conference on Security and Privacy in …, 2022 - Springer
We build our secure multiparty computation (MPC) protocols on top of the fully homomorphic
encryption (FHE) scheme, BFVrns, and augment it with Proxy Re-Encryption (PRE). We offer …

Secure over-the-air vehicle updates using trusted execution environments (TEE)

ACP Henriques - 2022 - search.proquest.com
Abstract Over-the-air (OTA) software update system has emerged as an important feature to
remotely analyze and upgrade the vehicle inside systems, coordinated by different ECUs …

[PDF][PDF] Rapid Deployment of Confidential Cloud Applications with Gramine

M Kowalczyk, D Kuvaiskii, P Marczewski, B Popławski… - cs.unc.edu
Gramine is a leading open-source tool for securely porting unmodified Linux applications
onto Intel® SGX [67]. Gramine implements the “lift-and-shift” model of confidential computing …