Existing federated learning solutions focus on transmitting features, parameters or gadients between clients and server, which suffer from serious low-efficiency and privacy-leakage problems. Thanks to the emerging foundation generative models, we propose a novel federated learning framework, namely Federated Generative Learning, that transmits prompts associated with distributed training data between clients and server. The informative training data can be synthesized remotely based on received prompts containing little privacy and the foundation generative models. The new framework possesses multiple advantages, including improved communication efficiency, better resilience to distribution shift, substantial performance gains, and enhanced privacy protection, which are verified in extensive experiments on ImageNet and DomainNet datasets.

翻译：现有联邦学习解决方案主要通过在客户端与服务器之间传输特征、参数或梯度，存在严重的低效率与隐私泄露问题。得益于新兴的基础生成式模型，我们提出了一种新型联邦学习框架——联邦生成式学习，该框架在客户端与服务器之间传输与分布式训练数据相关的提示。基于包含极少隐私信息的接收提示与基础生成式模型，可在远程端合成具有信息性的训练数据。该新框架具备多重优势，包括提升通信效率、增强对分布偏移的鲁棒性、显著性能增益以及强化隐私保护，这些优势在ImageNet与DomainNet数据集上的大量实验中得到验证。