FedDGHD: Diffusion Model-based Data Generation for Non-IID Challenges in Federated Learning

Federated Learning (FL) allows participants to collaboratively train a global model without sharing local data. However, the non-independent and identically distributed (non-lID) nature of data in real-world scenarios leads to a decline in the accuracy of existing federated learning methods. Using Generative Adversarial Networks (GANs) to generate realistic pseudo-data is a feasible approach to alleviate the non-lID issue. However, the realism of GAN-generated data comes at the cost of diversity and is prone to falling into local minima. Therefore, this paper innovatively integrates the pre-trained stable diffusion model's text-to-image mechanism with federated learning, constructing a client-server collaborative text generation mechanism that effectively alleviates the issue of data distribution bias. Furthermore, the effectiveness of FedDGHD is validated on datasets such as CIFARIO and CIFARIOO.