Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Nguyen-Le, Hong-Hanh, Tran, Van-Tuan, Nguyen, Thuc D., Le-Khac, Nhien-An
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2603.20304
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908903228309504
author Nguyen-Le, Hong-Hanh
Tran, Van-Tuan
Nguyen, Thuc D.
Le-Khac, Nhien-An
author_facet Nguyen-Le, Hong-Hanh
Tran, Van-Tuan
Nguyen, Thuc D.
Le-Khac, Nhien-An
contents As diffusion models (DMs) enable photorealistic image generation at unprecedented scale, watermarking techniques have become essential for provenance establishment and accountability. Existing methods face challenges: sampling-based approaches operate on frozen models but require costly $N$-step Denoising Diffusion Implicit Models (DDIM) inversion (typically N=50) for zero-bit-only detection; fine-tuning-based methods achieve fast multi-bit extraction but couple the watermark to a specific model checkpoint, requiring retraining for each architecture. We propose DiffMark, a plug-and-play watermarking method that offers three key advantages over existing approaches: single-pass multi-bit detection, per-image key flexibility, and cross-model transferability. Rather than encoding the watermark into the initial noise vector, DiffMark injects a persistent learned perturbation $δ$ at every denoising step of a completely frozen DM. The watermark signal accumulates in the final denoised latent $z_0$ and is recovered in a single forward pass. The central challenge of backpropagating gradients through a frozen UNet without traversing the full denoising chain is addressed by employing Latent Consistency Models (LCM) as a differentiable training bridge. This reduces the number of gradient steps from 50 DDIM to 4 LCM and enables a single-pass detection at 16.4 ms, a 45x speedup over sampling-based methods. Moreover, by this design, the encoder learns to map any runtime secret to a unique perturbation at inference time, providing genuine per-image key flexibility and transferability to unseen diffusion-based architectures without per-model fine-tuning. Although achieving these advantages, DiffMark also maintains competitive watermark robustness against distortion, regeneration, and adversarial attacks.
format Preprint
id arxiv_https___arxiv_org_abs_2603_20304
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Transferable Multi-Bit Watermarking Across Frozen Diffusion Models via Latent Consistency Bridges
Nguyen-Le, Hong-Hanh
Tran, Van-Tuan
Nguyen, Thuc D.
Le-Khac, Nhien-An
Computer Vision and Pattern Recognition
As diffusion models (DMs) enable photorealistic image generation at unprecedented scale, watermarking techniques have become essential for provenance establishment and accountability. Existing methods face challenges: sampling-based approaches operate on frozen models but require costly $N$-step Denoising Diffusion Implicit Models (DDIM) inversion (typically N=50) for zero-bit-only detection; fine-tuning-based methods achieve fast multi-bit extraction but couple the watermark to a specific model checkpoint, requiring retraining for each architecture. We propose DiffMark, a plug-and-play watermarking method that offers three key advantages over existing approaches: single-pass multi-bit detection, per-image key flexibility, and cross-model transferability. Rather than encoding the watermark into the initial noise vector, DiffMark injects a persistent learned perturbation $δ$ at every denoising step of a completely frozen DM. The watermark signal accumulates in the final denoised latent $z_0$ and is recovered in a single forward pass. The central challenge of backpropagating gradients through a frozen UNet without traversing the full denoising chain is addressed by employing Latent Consistency Models (LCM) as a differentiable training bridge. This reduces the number of gradient steps from 50 DDIM to 4 LCM and enables a single-pass detection at 16.4 ms, a 45x speedup over sampling-based methods. Moreover, by this design, the encoder learns to map any runtime secret to a unique perturbation at inference time, providing genuine per-image key flexibility and transferability to unseen diffusion-based architectures without per-model fine-tuning. Although achieving these advantages, DiffMark also maintains competitive watermark robustness against distortion, regeneration, and adversarial attacks.
title Transferable Multi-Bit Watermarking Across Frozen Diffusion Models via Latent Consistency Bridges
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.20304