OASIS: Offsetting Active Reconstruction Attacks in Federated Learning
Not on the map. Why?
Abstract (arXiv)
Federated Learning (FL) has garnered significant attention for its potential to protect user privacy while enhancing model training efficiency. For that reason, FL has found its use in various domains, from healthcare to industrial engineering, especially where data cannot be easily exchanged due to sensitive information or privacy laws. However, recent research has demonstrated that FL protocols can be easily compromised by active reconstruction attacks executed by dishonest servers. These attacks involve the malicious modification of global model parameters, allowing the server to obtain a verbatim copy of users' private data by inverting their gradient updates. Tackling this class of attack remains a crucial challenge due to the strong threat model. In this paper, we propose a defense mechanism, namely OASIS, based on image augmentation that effectively counteracts active reconstruction attacks while preserving model performance. We first uncover the core principle of gradient inversion that enables these attacks and theoretically identify the main conditions by which the defense can be robust regardless of the attack strategies. We then construct our defense with image augmentation showing that it can undermine the attack principle. Comprehensive evaluations demonstrate the efficacy of the defense mechanism highlighting its feasibility as a solution.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 24 of its 40 references that are in the map. First posted on arXiv 23 Nov 2023.
Builds on
24 of 40 references.
References · 24
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2016Deep Learning with Differential Privacycited by 2,728
- 2019Deep Leakage from Gradientscited by 1,021
- 2020Inverting Gradients -- How easy is it to break privacy in federated learning?cited by 712
- 2020Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixelscited by 585
- 2020iDLG: Improved Deep Leakage from Gradientscited by 382
- 2017Deep Models Under the GAN: Information Leakage from Collaborative Deep Learningcited by 334
- 2021See through Gradients: Image Batch Recovery via GradInversioncited by 326
- 2018Beyond Inferring Class Representatives: User-Level Privacy Leakage From Federated Learningcited by 246
- 2019Population Based Augmentation: Efficient Learning of Augmentation Policy Schedulescited by 154
- 2021When the Curious Abandon Honesty: Federated Learning Is Not Privatecited by 127
- 2020Multi-site fMRI Analysis Using Privacy-preserving Federated Learning and Domain Adaptation: ABIDE Resultscited by 122
12 more
- 2019Neural Blind Deconvolution Using Deep Priorscited by 113
- 2021Robbing the Fed: Directly Obtaining Private Data in Federated Learning with Modified Modelscited by 108
- 2020Provable Defense against Privacy Leakage in Federated Learning from Representation Perspectivecited by 94
- 2021Eluding Secure Aggregation in Federated Learning via Model Inconsistencycited by 67
- 2021HarmoFL: Harmonizing Local and Global Drifts in Federated Learning on Heterogeneous Medical Imagescited by 64
- 2018Federated Learning in Distributed Medical Databases: Meta-Analysis of Large-Scale Subcortical Brain Datacited by 46
- 2020Privacy-preserving Collaborative Learning with Automatic Transformation Searchcited by 31
- 2020Federated Learning in the Sky: Aerial-Ground Air Quality Sensing Framework with UAV Swarmscited by 26
- 2021Federated Learning on the Road: Autonomous Controller Design for Connected and Autonomous Vehiclescited by 26
- 2023Active Membership Inference Attack under Local Differential Privacy in Federated Learningcited by 17
- 2022Preserving Privacy and Security in Federated Learningcited by 13
- 2022Blockchain-based Secure Client Selection in Federated Learningcited by 5
Built on by
Not cited yet.