Browse > Article
http://dx.doi.org/10.13089/JKIISC.2021.31.5.987

Federated Learning Privacy Invasion Study in Batch Situation Using Gradient-Based Restoration Attack  

Jang, Jinhyeok (Soongsil University)
Ryu, Gwonsang (Soongsil University)
Choi, Daeseon (Soongsil University)
Publication Information
Journal of the Korea Institute of Information Security & Cryptology / v.31, no.5, 2021 , pp. 987-999 More about this Journal
Abstract
Recently, Federated learning has become an issue due to privacy invasion caused by data. Federated learning is safe from privacy violations because it does not need to be collected into a server and does not require learning data. As a result, studies on application methods for utilizing distributed devices and data are underway. However, Federated learning is no longer safe as research on the reconstruction attack to restore learning data from gradients transmitted in the Federated learning process progresses. This paper is to verify numerically and visually how well data reconstruction attacks work in various data situations. Considering that the attacker does not know how the data is constructed, divide the data with the class from when only one data exists to when multiple data are distributed within the class, and use MNIST data as an evaluation index that is MSE, LOSS, PSNR, and SSIM. The fact is that the more classes and data, the higher MSE, LOSS, and PSNR and SSIM are, the lower the reconstruction performance, but sufficient privacy invasion is possible with several reconstructed images.
Keywords
Data Privacy; Data Reconstruction; Federated learning;
Citations & Related Records
연도 인용수 순위
  • Reference
1 Z. Wang, A.C. Bovik, and H.R. Sheikh, "Image quality assessment: from error visibility to structural similarity", IEEE transactions on image processing, vol. 13, no. 4, pp. 600-612, Apr. 2004.   DOI
2 Python Pillow Library (Pillow - Pillow (PIL Fork)8.3.1 documentation ), Available:https://pillow.readthedocs.io/en/stable/, Accessed: Aug. 2021. [Online]
3 Google AI Blog, Federated Learning: Collaborative Machine Learning witho ut Centralized Training Data, Available:https://ai.googleblog.com/2017/04/federated-learning-collaborative.html, A ccessed: Jul. 2019. [Online]
4 Clement Fung, Chris J.M. Yoon, Ivan Beschastnikh, "Mitigating Sybils in Federated Learning Poisoning", arXiv, Jul. 2020.
5 Joungyoun Kim, Min-jeong Park, "Multiple imputation and synthetic data", The Korean Journal of Applied Statistics, vol. 32, no. 1, pp. 83-97, Feb. 2019.   DOI
6 Andrew Hard, Kanishka Rao, and Rajiv Mathews, "FEDERATED LEARNING FOR MOBILE KEYBOARD PREDICTION", arXiv, Feb. 2019.
7 H. Brendan McMahan, Eider Moore, and Daniel Ramage, "Communication-Efficient Learning of Deep Networks from Decentralized Data", Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, vol. 54, pp. 1273-1282, Feb. 2017.
8 Latanya Sweeney, "k-anonymity: a model for protecting privacy," International Journal on Uncertainty, Fuzziness and Knowledge-based Systems, vol. 10, no. 5, pp. 557-570, Jul. 2002.   DOI
9 Cynthia Dwork, Aaron Roth, "The Algorithmic Foundations of Differential Privacy", Foundations and Trends in Theoretical Computer Science, vol. 9, no. (3-4), pp. 211-407, Aug. 2014.   DOI
10 Max Jaderberg, Karen Simonyan, Andrea Vedaldi, and Andrew Zisserman, "Synthetic Data and Artificial Neural Networks for Natural Scene Text Recognition", arXiv, Dec. 2014.
11 Jooseok Park, "A Comparative Study of Big Data, Open Data, and My Data", The Korea Journal of BigData, vol. 3, no. 1, pp. 41-46, Aug. 2018.   DOI
12 Yue Zhao, Meng Li, Liangzhen Lai, and Naveen Suda, "Federated Learning with Non-IID Data", arXiv, Jun. 2018.
13 Tian Li, Anit Kumar Sahu, and Ameet Talwalkar, "Federated Learning: Challenges, methods, and future directions", IEEE SIGNAL PROCESSING MAGAZINE, vol. 37, no 3, pp. 50-60, May. 2020.   DOI
14 Keith Bonawitz, Hubert Eichner, and Wolfgang Grieskamp, "TOWARDS FEDERATED LEARNING AT SCALE: SYSTEM DESIGN", Proceedings of the 2nd SysML Conference, Mar. 2019.
15 Wenqi Wei, Ling Liu, Margaret Loper, and Ka-Ho Chow, "A Framework for Evaluating Clinet Privacy Leakages in Federated Learning", 25th European Symposium on Research in Computer Security, pp. 545-566, Sep. 2020.
16 H. Brendan, McMahan Eider, Moore Daniel Ramage et. al., "Communication-Efficient Learning of Deep Networks from Decentralized Data", Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS), vol. 54, pp. 1273-1282, Feb. 2017.
17 Vale Tolpegin, Stacey Truex, Mehmet Emre Gursoy, and Ling Liu, "Data Poisoning Attacks Against Federated Learning Systems", European Symposium on Research in Computer Security, pp. 480-501, Sep, 2020.
18 Eugene Bagdasaryan, Andreas Veit, and Yiqing Hua, "How To Backdoor Federated Learning", Proceedings of the 23rdInternational Conference on Artificial Intelligence and Statistics (AISTATS)2020, vol. 108, pp. 2938-2948, Aug. 2020.
19 Ligeng Zhu, Zhijian Liu, and Song Han, "Deep Leakage from Gradients", 33rd Conference on Neural Information Processing Systems NeurIPS, pp. 17-31, Dec. 2019.
20 Jonas Geiping, Hartmut Bauermeister, and Hannah Drog, "Inverting Gradients - How easy is it to break privacy in federated learning?", 34th Conference on Neural Information Processing Systems NeurIPS, Dec. 2020.
21 Qiang Yang, Yang Liu, and Tianjian Chen,"Federated Machine Learning: Concept and Applications", ACM Transactions on Intelligent Systems and Technology, vol. 10, no. 12, pp. 1-19, Jan. 2019.
22 Tian Li, Anit Kumar Sahu, and Ameet Talwalkar," Federated Learning: Challenges, methods, and future directions", IEEE SIGNAL PROCESSING MAGAZINE, vol. 37, no. 3, pp. 50-60, May. 2020.   DOI
23 Xin Yao, Tianchi Huang, and Chenglei Wu, "Federated Learning with Additional Mechanisms on Clients to Reduce Communication Costs", arXiv, Sep. 2019.
24 Tian Li, Anit Kumar Sahu, and Manzil Zaheer, "Federated Optimization in Heterogeneous Networks", Proceedings of the 3rd MLSys Conference, Apr. 2020.