FLClear: Visually Verifiable Multi-Client Watermarking for Federated Learning

2025-11-16

AI Summary

To address the vulnerability of client model intellectual property to server-side tampering or appropriation in federated learning, this paper proposes a multi-client verifiable watermarking framework. The method innovatively integrates transposed model architectures with contrastive learning to enable conflict-free watermark aggregation and enhanced security. It further introduces a dual-mode verification mechanism combining visual interpretability and structural similarity (SSIM) to support intuitive and robust ownership authentication. Extensive experiments across diverse datasets, aggregation strategies, and adversarial attack scenarios demonstrate that the proposed approach significantly outperforms state-of-the-art methods in watermark robustness, embedding fidelity, and verification reliability. Specifically, it achieves superior resilience against model pruning, fine-tuning, and watermark overwriting attacks while maintaining high accuracy in both watermark embedding and detection. The framework effectively ensures traceability and non-repudiation of client model contributions, thereby strengthening intellectual property protection in collaborative learning settings.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Computer Vision: Adversarial Attacks & Robustness
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Security and Privacy: Data transparency and provenance
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Federated learning (FL) enables multiple clients to collaboratively train a shared global model while preserving the privacy of their local data. Within this paradigm, the intellectual property rights (IPR) of client models are critical assets that must be protected. In practice, the central server responsible for maintaining the global model may maliciously manipulate the global model to erase client contributions or falsely claim sole ownership, thereby infringing on clients' IPR. Watermarking has emerged as a promising technique for asserting model ownership and protecting intellectual property. However, existing FL watermarking approaches remain limited, suffering from potential watermark collisions among clients, insufficient watermark security, and non-intuitive verification mechanisms. In this paper, we propose FLClear, a novel framework that simultaneously achieves collision-free watermark aggregation, enhanced watermark security, and visually interpretable ownership verification. Specifically, FLClear introduces a transposed model jointly optimized with contrastive learning to integrate the watermarking and main task objectives. During verification, the watermark is reconstructed from the transposed model and evaluated through both visual inspection and structural similarity metrics, enabling intuitive and quantitative ownership verification. Comprehensive experiments conducted over various datasets, aggregation schemes, and attack scenarios demonstrate the effectiveness of FLClear and confirm that it consistently outperforms state-of-the-art FL watermarking methods.

Problem

• Protecting client intellectual property rights in federated learning systems
• Preventing malicious server manipulation erasing client model contributions
• Addressing watermark collisions and non-intuitive verification in FL watermarking

Innovation

• Transposed model with contrastive learning integration
• Collision-free watermark aggregation for multiple clients
• Visual and structural similarity based verification