CoreMark: Toward Robust and Universal Text Watermarking Technique

2025-06-28

AI Summary

Existing text watermarking methods struggle to simultaneously achieve robustness, universality, and imperceptibility. This paper proposes CoreMark, a universal multilingual and multifont watermarking framework. Its core innovation lies in a novel embedding paradigm based on contiguous black pixel segments within characters—termed COREs—which inherently confers noise resilience. CoreMark further incorporates four key modules: dynamic CORE extraction, length-driven robust character selection, CORE thickness modulation, and font-adaptive intensity adjustment. Experiments demonstrate that CoreMark significantly outperforms state-of-the-art methods under severe distortions—including screenshots, print-scan cycles, and smartphone photography—while preserving visual fidelity. Moreover, it exhibits high universality and stability across over 20 languages and 50+ fonts, validating its broad applicability and practical viability.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Natural Language Processing: Safety and Robustness

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Responsible Web: Measurement, analysis, and circumvention of Web censorship
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Text watermarking schemes have gained considerable attention in recent years, yet still face critical challenges in achieving simultaneous robustness, generalizability, and imperceptibility. This paper introduces a new embedding paradigm,termed CORE, which comprises several consecutively aligned black pixel segments. Its key innovation lies in its inherent noise resistance during transmission and broad applicability across languages and fonts. Based on the CORE, we present a text watermarking framework named CoreMark. Specifically, CoreMark first dynamically extracts COREs from characters. Then, the characters with stronger robustness are selected according to the lengths of COREs. By modifying the thickness of the CORE, the hidden data is embedded into the selected characters without causing significant visual distortions. Moreover, a general plug-and-play embedding strength modulator is proposed, which can adaptively enhance the robustness for small font sizes by adjusting the embedding strength according to the font size. Experimental evaluation indicates that CoreMark demonstrates outstanding generalizability across multiple languages and fonts. Compared to existing methods, CoreMark achieves significant improvements in resisting screenshot, print-scan, and print camera attacks, while maintaining satisfactory imperceptibility.

Problem

• Achieving robust, generalizable, and imperceptible text watermarking
• Resisting noise during transmission across languages and fonts
• Enhancing robustness for small fonts via adaptive strength modulation

Innovation

• CORE segments enhance noise resistance
• Dynamic CORE extraction and selection
• Adaptive strength modulator for fonts