ALIEN: Analytic Latent Watermarking for Controllable Generation

2026-02-05

AI Summary

Existing latent watermarking methods rely on computationally intensive heuristic optimization, struggling to balance robustness and generation quality. This work proposes ALIEN, a novel framework that, for the first time, enables analytical control over the watermark embedding process. By deriving time-varying modulation coefficients, ALIEN precisely steers the propagation of watermark residuals throughout the diffusion trajectory—eliminating the need for iterative optimization. Built upon latent diffusion models, ALIEN achieves substantial gains in both efficiency and performance: its variant ALIEN-Q surpasses state-of-the-art methods by an average of 33.1% across five generation quality metrics, while ALIEN-R demonstrates a 14.0% improvement in robustness under 15 diverse perturbations.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Learning to Search

Application Category

• Responsible Web: Algorithmic accountability and transparency on the web
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Web Mining and Content Analysis: Web data quality in the era of algorithmically-generated content

Abstract

Watermarking is a technical alternative to safeguarding intellectual property and reducing misuse. Existing methods focus on optimizing watermarked latent variables to balance watermark robustness and fidelity, as Latent diffusion models (LDMs) are considered a powerful tool for generative tasks. However, reliance on computationally intensive heuristic optimization for iterative signal refinement results in high training overhead and local optima entrapment.To address these issues, we propose an \underline{A}na\underline{l}ytical Watermark\underline{i}ng Framework for Controllabl\underline{e} Generatio\underline{n} (ALIEN). We develop the first analytical derivation of the time-dependent modulation coefficient that guides the diffusion of watermark residuals to achieve controllable watermark embedding pattern.Experimental results show that ALIEN-Q outperforms the state-of-the-art by 33.1\% across 5 quality metrics, and ALIEN-R demonstrates 14.0\% improved robustness against generative variant and stability threats compared to the state-of-the-art across 15 distinct conditions. Code can be available at https://anonymous.4open.science/r/ALIEN/.

Problem

• watermarking
• latent diffusion models
• heuristic optimization
• robustness
• fidelity

Innovation

• Analytic Watermarking
• Latent Diffusion Models
• Time-dependent Modulation
• Controllable Generation
• Watermark Robustness