analiza aiqbench
This research addresses a critical gap in robust AI by applying adversarial training to vision transformers within federated learning environments, where Non-IID data distributions often degrade model performance. The authors introduce FedWAvg, an enhanced aggregation technique that leverages similarity metrics between global and local model layers to improve robust accuracy against evasion attacks. This work advances the intersection of privacy-preserving machine learning and defensive AI strategies for transformer architectures.
Federated Adversarial Training with Transformers
arXiv:2206.02131v2 Announce Type: replace Abstract: Federated learning (FL) has emerged to enable global model training over distributed clients' data while preserving its privacy. However, the global trained model is vulnerable to the evasion attacks especially, the adversarial examples (AEs), carefully crafted samples to yield false classification. Adversarial training (AT) is found to be the most promising approach against evasion attacks and it is widely studied for convolutional neural network (CNN). Recently, vision transformers have been found to be effective in many computer vision tasks. To the best of the authors' knowledge, there is no work that studied the feasibility of AT in a FL process for vision transformers. This paper investigates such feasibility with different federated model aggregation methods and different vision transformer models with different tokenization and classification head techniques. In order to improve the robust accuracy of the models with the not independent and identically distributed (Non-IID), we propose an extension to FedAvg aggregation method, called FedWAvg. By measuring the similarities between the last layer of the global model and the last layer of the client updates, FedWAvg calculates the weights to aggregate the local models updates. The experiments show that FedWAvg improves the robust accuracy when compared with other state-of-the-art aggregation methods.
Abstract:Federated learning (FL) has emerged to enable global model training over distributed clients' data while preserving its privacy. However, the global trained model is vulnerable to the evasion attacks especially, the adversarial examples (AEs), carefully crafted samples to yield false classification. Adversarial training (AT) is found to be the most promising approach against evasion attacks and it is widely studied for convolutional neural network (CNN). Recently, vision transformers have been found to be effective in many computer vision tasks. To the best of the authors' knowledge, there is no work that studied the feasibility of AT in a FL process for vision transformers. This paper investigates such feasibility with different federated model aggregation methods and different vision transformer models with different tokenization and classification head techniques. In order to improve the robust accuracy of the models with the not independent and identically distributed (Non-IID), we propose an extension to FedAvg aggregation method, called FedWAvg. By measuring the similarities between the last layer of the global model and the last layer of the client updates, FedWAvg calculates the weights to aggregate the local models updates. The experiments show that FedWAvg improves the robust accuracy when compared with other state-of-the-art aggregation methods.
Machine Learning (cs.LG); Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)
arXiv:2206.02131 [cs.LG]
(or arXiv:2206.02131v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2206.02131 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 302 cuvinte
Subiecte conexe
[AINews] Reality Checks on AI News (Yegge shuts down Gas Town, Databricks’ +60% Astra cost)
Latent Space · 17 sept. 2026
Novel Methods for Catheter and Guidewire Segmentation in X-ray Fluoroscopy under a Federated Learning Setting
arXiv · 9 sept. 2026
Cooperative Multi-Agent Reinforcement Learning for Adaptive Aggregation in Semi-Supervised Federated Learning with non-IID Data
arXiv · 27 aug. 2026
FLINT: Fingerprinting Federated Learning Architectures from 5G PHY-Layer Side Channels
arXiv · 20 iul. 2026