Semantic Watermarking Reinvented: Enhancing Robustness and Generation Quality with Fourier Integrity

2025-09-09

AI Summary

Semantic watermarking in latent diffusion models (LDMs) faces two key challenges: degradation in detection performance due to frequency-domain integrity loss and poor robustness against cropping. To address these, we propose Symmetric Fourier Watermarking (SFW), a novel Hermitian-symmetric frequency-domain watermarking framework. SFW is the first to impose Hermitian symmetry constraints on the Fourier spectrum to preserve structural consistency in the frequency domain; it introduces a center-aware embedding mechanism to enhance cropping resilience; and it synergistically integrates semantic watermarking for joint frequency-domain and semantic optimization. The method maintains superior image fidelity—achieving state-of-the-art FID and CLIP Score—and significantly improves robustness against both model regeneration and arbitrary cropping. Under diverse adversarial attacks, SFW attains optimal detection accuracy and identification robustness, effectively balancing watermark robustness, generative quality, and information capacity.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Safety and Robustness

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Semantic watermarking techniques for latent diffusion models (LDMs) are robust against regeneration attacks, but often suffer from detection performance degradation due to the loss of frequency integrity. To tackle this problem, we propose a novel embedding method called Hermitian Symmetric Fourier Watermarking (SFW), which maintains frequency integrity by enforcing Hermitian symmetry. Additionally, we introduce a center-aware embedding strategy that reduces the vulnerability of semantic watermarking due to cropping attacks by ensuring robust information retention. To validate our approach, we apply these techniques to existing semantic watermarking schemes, enhancing their frequency-domain structures for better robustness and retrieval accuracy. Extensive experiments demonstrate that our methods achieve state-of-the-art verification and identification performance, surpassing previous approaches across various attack scenarios. Ablation studies confirm the impact of SFW on detection capabilities, the effectiveness of the center-aware embedding against cropping, and how message capacity influences identification accuracy. Notably, our method achieves the highest detection accuracy while maintaining superior image fidelity, as evidenced by FID and CLIP scores. Conclusively, our proposed SFW is shown to be an effective framework for balancing robustness and image fidelity, addressing the inherent trade-offs in semantic watermarking. Code available at https://github.com/thomas11809/SFWMark

Problem

• Enhancing robustness against regeneration and cropping attacks
• Maintaining frequency integrity in semantic watermarking techniques
• Balancing detection performance with superior image generation quality

Innovation

• Hermitian symmetric Fourier watermarking for integrity
• Center-aware embedding strategy against cropping attacks
• Enhanced frequency-domain structures for robustness