arXiv cs.AI· Yongcun Song, Ziqi Wang, Enrique Zuazua·· 1 天前AI 评分31
联邦学习中的近似加权数据重建攻击（AWA）
Approximate and Weighted Data Reconstruction Attack in Federated Learning
AI 导读
针对现有数据重建攻击难以攻破横向 FedAvg 场景的问题，研究者提出近似加权攻击方法 AWA：用基于插值的近似方法生成客户端本地训练的中间模型更新，使 FedAvg 场景可被攻击，并设计逐层加权损失函数、由贝叶斯优化调参以提升重建质量。实验显示 AWA 在图像数据重建的多项评估指标上显著优于现有 SOTA 方法，代码已开源。
正文 · 原文
Abstract:Federated Learning (FL) is a distributed learning paradigm that enables multiple clients to collaborate on building a machine learning model without sharing their private data. Although FL is considered privacy-preserved by design, recent data reconstruction attacks demonstrate that an attacker can recover clients' training data based on the parameters shared in FL. However, most existing methods fail to attack the most widely used horizontal Federated Averaging (FedAvg) scenario, where clients share model parameters after multiple local training steps. To tackle this issue, we propose an interpolation-based approximation method, which makes attacking FedAvg scenarios feasible by generating the intermediate model updates of the clients' local training processes. Then, we design a layer-wise weighted loss function to improve the data quality of reconstruction. We assign different weights to model updates in different layers concerning the neural network structure, with the weights tuned by Bayesian optimization. Finally, experimental results validate the superiority of our proposed approximate and weighted attack (AWA) method over the other state-of-the-art methods, as demonstrated by the substantial improvement in different evaluation metrics for image data reconstructions. The source code is available at this https URL.
| Comments: | Funded by the European Union (HORIZON-MSCA-2021-DN-01-01 project ModConFlex, grant number 101073558). Code available at this https URL. Code DOI: this https URL |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Optimization and Control (math.OC) |
| Cite as: | arXiv:2308.06822 [cs.LG] |
| (or arXiv:2308.06822v4 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2308.06822 arXiv-issued DOI via DataCite |
|
| Journal reference: | IEEE Transactions on Big Data, vol. 12, no. 5, pp. 1606-1617, 2026 |
| Related DOI: | https://doi.org/10.1109/TBDATA.2026.3695416
DOI(s) linking to related resources |
Submission history
From: Ziqi Wang [view email]
[v1]
Sun, 13 Aug 2023 17:40:56 UTC (1,859 KB)
[v2]
Tue, 26 Mar 2024 19:39:23 UTC (1,384 KB)
[v3]
Fri, 15 May 2026 12:55:02 UTC (6,844 KB)
[v4]
Thu, 12 Nov 2026 16:04:54 UTC (6,842 KB)
来源：arXiv cs.AI · arxiv.org