跳到正文
arXiv cs.AI· Yixuan Tong, Yuanguo Bi, Mingjian Zhi, Wenchao Xu, Yuan Wu·· 1 小时前AI 评分35

FedDoG:基于服务器端预训练扩散模型的一次性联邦学习免训练域生成方法

Training-Free Domain Generation for One-Shot Federated Learning with Server-Side Pre-trained Diffusion Models

AI 导读

研究者提出 FedDoG,一个免训练的域生成框架,利用服务器端预训练扩散模型为一次性联邦学习(OSFL)合成对齐客户端域的数据。该方法用冻结图像编码器提取紧凑域表示,并通过预训练图像条件适配器注入扩散模型,同时根据域集中度调节视觉条件强度、按类一致性与预测熵过滤样本。在三个基准上,FedDoG 的分类准确率超过现有 OSFL 基线,客户端上传和计算时间分别减少 99.9% 和 91.5%。

正文 · 原文

View PDF HTML (experimental)

Abstract:One-shot federated learning (OSFL) has attracted increasing attention for its ability to substantially reduce communication overhead by limiting client-server interaction to a single communication round. Nevertheless, this limited interaction makes learning from heterogeneous client distributions challenging. Diffusion-based OSFL approaches seek to address this challenge by synthesizing data that approximate client distributions. However, existing diffusion-based approaches often impose substantial computational costs on clients and lack effective visual guidance to generate data aligned with client distributions. To address these limitations, we propose FedDoG, a training-free Domain Generation framework that leverages server-side pre-trained diffusion models to synthesize data aligned with client domains. FedDoG extracts compact domain representations using a frozen image encoder, substantially reducing client-side computational and communication costs. These representations are incorporated into the diffusion model through a pre-trained image-conditioning adapter, thereby providing explicit visual guidance for domain generation. To further improve the quality of synthetic data, FedDoG adapts the strength of visual conditioning according to domain concentration and filters generated samples based on class consistency and predictive entropy. Experiments on three benchmarks demonstrate that FedDoG generates higher-quality synthetic images and outperforms existing OSFL baselines in classification accuracy. Meanwhile, FedDoG reduces client upload and computation time by 99.9% and 91.5%, respectively. The code is available at this https URL.
Subjects: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Distributed, Parallel, and Cluster Computing (cs.DC)
Cite as: arXiv:2611.23908 [cs.AI]
  (or arXiv:2611.23908v1 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2611.23908

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Yixuan Tong [view email]
[v1] Wed, 18 Nov 2026 05:59:45 UTC (1,848 KB)

来源:arXiv cs.AI · arxiv.org