FedOT: Ownership Verification and Leakage Tracing via Watermarks for Federated LDMs

Hugging Face Daily Papers Papers
federated-learning latent-diffusion-models watermarking ownership-verification leakage-tracing security ai-safety

Summary

FedOT introduces a chunked watermarking and latent vector transformation framework for ownership verification and leakage tracing in federated latent diffusion models, preventing watermark removal attacks.

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/.
Original Article
View Cached Full Text

Cached at: 06/24/26, 05:46 AM

Paper page - FedOT: Ownership Verification and Leakage Tracing via Watermarks for Federated LDMs

Source: https://huggingface.co/papers/2606.22875

Abstract

FedOT is a novel framework that enables ownership verification and leakage tracing in federated latent diffusion models by introducing chunked watermarking and latent vector transformation to prevent watermark removal attacks.

TrainingLatent Diffusion Models(LDMs) withinFederated 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 existingVAE-based watermarkingtechniques for LDMs in FL, this strategy falls short in addressing such threats due to two fundamental challenges: (1) Existing methods supportownership verificationbut 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 forownership verificationandleakage tracinginfederated LDMs. Specifically, to address the first challenge, we design a chunked watermark, where the first part is forownership 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 introduceLatent Vector Transformation(LVT), which strengthens the connection between the VAE andU-Netlatent 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 bothownership verificationand traceability. Project page: https://spyzixuan.github.io/FedOT/.

View arXiv pageView PDFProject pageAdd to collection

Get this paper in your agent:

hf papers read 2606\.22875

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.22875 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.22875 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.22875 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Dataset Watermarking for Closed LLMs with Provable Detection

arXiv cs.LG

This paper introduces a novel dataset watermarking method for closed LLMs that uses co-occurrence patterns of word pairs to provably detect if proprietary data was used in training, even when it constitutes a small fraction of the dataset.

Robust Text Watermarking for Large Language Models via Dual Semantic Embeddings

arXiv cs.CL

This paper presents Dual-Embedding Watermarking (DEW), a semantic watermarking scheme for LLMs that improves robustness against paraphrasing and translation by leveraging contextual and token-level embeddings. Experimental results show improved detection after paraphrasing and translation compared to prior methods.

Linguistics-Aware Non-Distortionary LLM Watermarking

arXiv cs.CL

Introduces LUNA, a linguistics-aware LLM watermarking method that achieves non-distortionary embedding and model-free detection across multiple languages, significantly improving AUROC and perplexity preservation.