Collaborative Synthetic Data Generation for Knowledge Transfer in Federated Learning

2026-07-08

AI Summary

This work addresses the challenge in single-round federated learning of simultaneously achieving data heterogeneity robustness, low communication overhead, and strong privacy guarantees. The authors propose a novel approach wherein clients leverage a publicly available pre-trained neural compression autoencoder to encode their private data into a shared latent space and upload class-conditional statistics. The server aggregates these statistics via secure aggregation, injects $(\varepsilon, \delta)$-differential privacy noise, and decodes the perturbed aggregate to generate synthetic data for global model training. This method uniquely integrates pre-trained autoencoders, secure aggregation, and differential privacy to enable efficient knowledge transfer under rigorous privacy constraints. Experiments demonstrate that the approach matches or even surpasses non-private baselines across diverse datasets and heterogeneous settings, while exhibiting strong scalability to large numbers of clients.

Abstract

One-shot federated learning (OSFL) addresses the communication overhead of federated learning by limiting training to a single round, but doing so without sacrificing model quality is non-trivial, particularly when client data distributions diverge. Recent work has addressed this challenge by aggregating client knowledge on the server through the construction of transferable synthetic datasets or distillates. However, most of these methods lack formal privacy guarantees, leaving a gap in jointly achieving low communication, robustness to heterogeneity, and rigorous privacy. We propose FedKT-CSD (Federated Knowledge Transfer via Collaborative Synthetic Data), a framework inspired by neural image compression that closes this gap by leveraging publicly pretrained autoencoders as a shared latent space. Each client encodes its private data in a single forward pass, computes class-conditional latent statistics, and transmits these to the server. The server aggregates these statistics via secure aggregation, adds calibrated differential privacy noise, and decodes a synthetic dataset for training a global model and further downstream tasks. This design provides formal $(\varepsilon,δ)$-differential privacy by construction, while keeping client-side computation and communication lightweight. Despite operating under privacy constraints, FedKT-CSD is competitive with and even outperforms non-private baselines across diverse datasets and heterogeneity settings, and scales to a large number of clients. Our code is available at: https://github.com/an7123/FedKT-CSD

Problem

• Federated Learning
• One-shot Learning
• Differential Privacy
• Data Heterogeneity
• Communication Efficiency

Innovation

• Federated Learning
• Differential Privacy
• Synthetic Data Generation
• One-shot Learning
• Knowledge Transfer