Gaussian Shannon: High-Precision Diffusion Model Watermarking Based on Communication

TL;DR

Gaussian Shannon introduces a watermarking framework for diffusion models that enables lossless, bit-exact watermark recovery by modeling the process as a noisy communication channel and employing error correction.

Contribution

It proposes a novel watermarking method that embeds watermarks in initial noise and ensures exact bit recovery without fine-tuning or quality loss.

Findings

Achieves state-of-the-art bit-level accuracy across multiple diffusion models.

Maintains high true positive rates for watermark detection.

Effectively resists local and global distortions in real-world scenarios.

Abstract

Diffusion models generate high-quality images but pose serious risks like copyright violation and disinformation. Watermarking is a key defense for tracing and authenticating AI-generated content. However, existing methods rely on threshold-based detection, which only supports fuzzy matching and cannot recover structured watermark data bit-exactly, making them unsuitable for offline verification or applications requiring lossless metadata (e.g., licensing instructions). To address this problem, in this paper, we propose Gaussian Shannon, a watermarking framework that treats the diffusion process as a noisy communication channel and enables both robust tracing and exact bit recovery. Our method embeds watermarks in the initial Gaussian noise without fine-tuning or quality loss. We identify two types of channel interference, namely local bit flips and global stochastic distortions, and design a cascaded defense combining error-correcting codes and majority voting. This ensures reliable end-to-end transmission of semantic payloads. Experiments across three Stable Diffusion variants and seven perturbation types show that Gaussian Shannon achieves state-of-the-art bit-level accuracy while maintaining a high true positive rate, enabling trustworthy rights attribution in real-world deployment. The source code have been made available at: https://github.com/Rambo-Yi/Gaussian-Shannon