arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2312.02247cs.LGcs.AIcs.CV
收藏

面向目标域泛化的联邦主动学习

Federated Active Learning for Target Domain Generalisation

University College London(伦敦大学学院);University of Aberdeen(阿伯丁大学)

Razvan Caramalau, Binod Bhattarai, Danail Stoyanov

AI总结:

提出融合联邦域泛化与可变预算主动学习的 FEDALV 框架,通过客户端、服务器双重优化和信息性数据采样,仅用 5% 源客户端数据即达到完整训练的目标准确率。

摘要:

本文将主动学习框架引入联邦学习,以实现目标域泛化,从而结合两种学习范式的优势。我们的框架 FEDALV 由主动学习(AL)和联邦域泛化(FDG)组成,能够在不共享图像的情况下,将利用有限源域客户端数据训练的图像分类模型泛化到未见过的目标域。为此,我们的 FDG 方法 FEDA 在训练过程中包含两次优化更新:一次在客户端,另一次在服务器端。在客户端,所引入的损失旨在降低特征复杂性并进行条件对齐;而在服务器端,正则化限制全局模型获得的源域与目标域之间的自由能偏差。FEDAL 的其余组成部分是具有可变预算的 AL,它向服务器查询,以为目标客户端检索并采样信息量最大的本地数据。我们在带 AL 和不带 AL 的 FDG 上进行了多项实验,并与传统 FDG 基线和联邦主动学习基线进行了比较。大量定量实验表明,与多种同期方法相比,我们的方法在准确率和效率方面具有优越性。FEDALV 仅采样源客户端少至 5% 的数据,即可达到完整训练目标准确率的性能。

英文摘要:

In this paper, we introduce Active Learning framework in Federated Learning for Target Domain Generalisation, harnessing the strength from both learning paradigms. Our framework, FEDALV, composed of Active Learning (AL) and Federated Domain Generalisation (FDG), enables generalisation of an image classification model trained from limited source domain client's data without sharing images to an unseen target domain. To this end, our FDG, FEDA, consists of two optimisation updates during training, one at the client and another at the server level. For the client, the introduced losses aim to reduce feature complexity and condition alignment, while in the server, the regularisation limits free energy biases between source and target obtained by the global model. The remaining component of FEDAL is AL with variable budgets, which queries the server to retrieve and sample the most informative local data for the targeted client. We performed multiple experiments on FDG w/ and w/o AL and compared with both conventional FDG baselines and Federated Active Learning baselines. Our extensive quantitative experiments demonstrate the superiority of our method in accuracy and efficiency compared to the multiple contemporary methods. FEDALV manages to obtain the performance of the full training target accuracy while sampling as little as 5% of the source client's data.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑