Enabling Execution Assurance of Federated Learning at Untrusted Participants

Zhang, Xiaoli; Li, Fengting; Zhang, Zeyu; Li, Qi; Wu, Jianping

doi:10.1109/infocom41043.2020.9155414

Cited by 68 publications

(22 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…DarkneTZ's evaluation showed no more than 10% overhead in CPU, memory, and energy on edge-like devices, demonstrating its suitability for client-side model updates in FL. In an orthogonal direction, several works leveraged clients' TEEs for verifying the integrity of local model training [10,75], but did not consider privacy. Considering a broader range of attacks (e.g., DRAs and PIAs), it is essential to protect all layers instead of the last layers only, something that PPFL does.…”

Section: Privacy-preserving ML Using Teesmentioning

confidence: 99%

PPFL

Haddadi

Katevas

et al. 2021

Proceedings of the 19th Annual International Conference on Mobile Systems, Applications, and Services

144

View full text Add to dashboard Cite

We propose and implement a Privacy-preserving Federated Learning () framework for mobile systems to limit privacy leakages in federated learning. Leveraging the widespread presence of Trusted Execution Environments (TEEs) in high-end and mobile devices, we utilize TEEs on clients for local training, and on servers for secure aggregation, so that model/gradient updates are hidden from adversaries. Challenged by the limited memory size of current TEEs, we leverage greedy layer-wise training to train each model's layer inside the trusted area until its convergence. The performance evaluation of our implementation shows that can significantly improve privacy while incurring small system overheads at the client-side. In particular, can successfully defend the trained model against data reconstruction, property inference, and membership inference attacks. Furthermore, it can achieve comparable model utility with fewer communication rounds (0.54×) and a similar amount of network traffic (1.002×) compared to the standard federated learning of a complete model. This is achieved while only introducing up to ∼15% CPU time, ∼18% memory usage, and ∼21% energy consumption overhead in 's client-side. CCS CONCEPTS• Security and privacy → Privacy protections; Distributed systems security; • Computing methodologies → Distributed algorithms.

show abstract

Section: Privacy-preserving ML Using Teesmentioning

confidence: 99%

PPFL

Haddadi

Katevas

et al. 2021

Proceedings of the 19th Annual International Conference on Mobile Systems, Applications, and Services

144

View full text Add to dashboard Cite

show abstract

“…We will briefly review recent related work in two categories: (i) studies either federated learning (FedL) (e.g., [5,19,20,23,43,46]) or few-shot learning (FSL) (e.g., [9,16,22,24,[34][35][36]38]), or both of them [3] (ii) studies proposing similar ideas of minimizing model divergence to better learn individual models or an ensemble model.…”

Section: Related Workmentioning

confidence: 99%

Federated Few-Shot Learning with Adversarial Learning

Cheng

Huang

2021

Preprint

View full text Add to dashboard Cite

We are interested in developing a unified machine learning model over many mobile devices for practical learning tasks, where each device only has very few training data. This is a commonly encountered situation in mobile computing scenarios, where data is scarce and distributed while the tasks are distinct. In this paper, we propose a federated few-shot learning (FedFSL) framework to learn a fewshot classification model that can classify unseen data classes with only a few labeled samples. With the federated learning strategy, FedFSL can utilize many data sources while keeping data privacy and communication efficiency. There are two technical challenges: 1) directly using the existing federated learning approach may lead to misaligned decision boundaries produced by client models, and 2) constraining the decision boundaries to be similar over clients would overfit to training tasks but not adapt well to unseen tasks. To address these issues, we propose to regularize local updates by minimizing the divergence of client models. We also formulate the training in an adversarial fashion and optimize the client models to produce a discriminative feature space that can better represent unseen data samples. We demonstrate the intuitions and conduct experiments to show our approaches outperform baselines by more than 10% in learning vision tasks and 5% in language tasks. CCS CONCEPTS• Computing methodologies → Artificial intelligence; Distributed artificial intelligence; Computer vision.

show abstract

“…They proposed a control algorithm that determines the best frequency of global aggregation with which computation and communication resources at the edge can be used efficiently in federated learning. Zhang et al [27] proposed building trustworthy federated learning systems using trusted execution environments (TEEs). Their main focus was to assure that the local training on clients side is being done correctly.…”

Section: Related Workmentioning

confidence: 99%

Privacy-Preserving Learning of Human Activity Predictors in Smart Environments

Zehtabian¹,

Khodadadeh²,

Bölöni³

et al. 2021

Preprint

View full text Add to dashboard Cite

The daily activities performed by a disabled or elderly person can be monitored by a smart environment, and the acquired data can be used to learn a predictive model of user behavior. To speed up the learning, several researchers designed collaborative learning systems that use data from multiple users. However, disclosing the daily activities of an elderly or disabled user raises privacy concerns.In this paper, we use state-of-the-art deep neural networkbased techniques to learn predictive human activity models in the local, centralized, and federated learning settings. A novel aspect of our work is that we carefully track the temporal evolution of the data available to the learner and the data shared by the user. In contrast to previous work where users shared all their data with the centralized learner, we consider users that aim to preserve their privacy. Thus, they choose between approaches in order to achieve their goals of predictive accuracy while minimizing the shared data. To help users make decisions before disclosing any data, we use machine learning to predict the degree to which a user would benefit from collaborative learning. We validate our approaches on real-world data.

show abstract

Enabling Execution Assurance of Federated Learning at Untrusted Participants

Cited by 68 publications

References 20 publications

PPFL

PPFL

Federated Few-Shot Learning with Adversarial Learning

Privacy-Preserving Learning of Human Activity Predictors in Smart Environments

Contact Info

Product

Resources

About