AI summaryⓘ
The authors address the problem of protecting the ownership and tracing leaks of Latent Diffusion Models (LDMs) trained via Federated Learning (FL), where models are shared across multiple clients. They point out that existing watermarking methods only prove ownership but cannot identify which client leaked the model, and these methods are vulnerable to attacks that remove watermarks by swapping model parts. To fix this, they propose FedOT, which uses a two-part watermark to both verify ownership and trace the leaking client, and introduce a technique called Latent Vector Transformation to tie components of the model together more tightly, preventing watermark removal without ruining model quality. Their experiments show that FedOT effectively ensures ownership verification and can trace leaks in federated LDMs.
Latent Diffusion ModelsFederated LearningWatermarkingOwnership VerificationLeakage TracingVariational AutoencoderU-NetLatent Vector TransformationModel SecurityDecoder Replacement Attack
Authors
Wenlong Cheng, Yuan Gan, Yunqiu Xu, Jiaxu Miao
Abstract
Training Latent Diffusion Models (LDMs) within Federated Learning (FL) has attracted increasing attention due to its ability to combine the powerful generative capacity of LDMs with the privacy-preserving properties of FL. However, FL requires sharing the global model with multiple participants, which risks unauthorized model distribution or resale by malicious clients. While an intuitive approach is to adopt existing VAE-based watermarking techniques for LDMs in FL, this strategy falls short in addressing such threats due to two fundamental challenges: (1) Existing methods support ownership verification but lack the ability to trace model leakage to a specific malicious client; (2) VAE-based watermarks are vulnerable, as they can be removed simply by replacing the decoder with a clean counterpart. In this paper, we propose FedOT, the first framework for ownership verification and leakage tracing in federated LDMs. Specifically, to address the first challenge, we design a chunked watermark, where the first part is for ownership verification, and the second part is used for client identification. Furthermore, to overcome the second challenge and secure the model against VAE replacement attack, we introduce Latent Vector Transformation (LVT), which strengthens the connection between the VAE and U-Net latent spaces by modifying the original latent distribution of the VAE. Consequently, any attempt to replace the VAE for watermark removal leads to significant image quality degradation, making the LDM model unusable. Extensive experiments demonstrate that FedOT achieves superior performance in both ownership verification and traceability. Project page: https://spyzixuan.github.io/FedOT/.