Learning to Watermark in the Latent Space of Generative Models

• URL: http://arxiv.org/abs/2601.16140v1
• Date: Thu, 22 Jan 2026 17:34:30 GMT
• Title: Learning to Watermark in the Latent Space of Generative Models
• Authors: Sylvestre-Alvise Rebuffi, Tuan Tran, Valeriu Lacatusu, Pierre Fernandez, Tomáš Souček, Nikola Jovanović, Tom Sander, Hady Elsahar, Alexandre Mourachko,
• Abstract summary: DistSeal is a unified approach for latent watermarking that works across both diffusion and autoregressive models.<n>Our approach works by training post-hoc watermarking models in the latent space of generative models.
• Score: 42.29703361234457
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Existing approaches for watermarking AI-generated images often rely on post-hoc methods applied in pixel space, introducing computational overhead and potential visual artifacts. In this work, we explore latent space watermarking and introduce DistSeal, a unified approach for latent watermarking that works across both diffusion and autoregressive models. Our approach works by training post-hoc watermarking models in the latent space of generative models. We demonstrate that these latent watermarkers can be effectively distilled either into the generative model itself or into the latent decoder, enabling in-model watermarking. The resulting latent watermarks achieve competitive robustness while offering similar imperceptibility and up to 20x speedup compared to pixel-space baselines. Our experiments further reveal that distilling latent watermarkers outperforms distilling pixel-space ones, providing a solution that is both more efficient and more robust.

Related papers

• Optimization-Free Universal Watermark Forgery with Regenerative Diffusion Models [50.73220224678009]
  Watermarking can be used to verify the origin of synthetic images generated by artificial intelligence models.<n>Recent studies demonstrate the capability to forge watermarks from a target image onto cover images via adversarial techniques.<n>In this paper, we uncover a greater risk of an optimization-free and universal watermark forgery.<n>Our approach significantly broadens the scope of attacks, presenting a greater challenge to the security of current watermarking techniques.
  arXiv  Detail & Related papers  (2025-06-06T12:08:02Z)
• Image Watermarking of Generative Diffusion Models [42.982489491857145]
  We propose a watermarking technique that embeds watermark features into the diffusion model itself.<n>Our technique enables training of a paired watermark extractor for a generative model that is learned through an end-to-end process.<n>We demonstrate highly accurate watermark embedding/detection and show that it is also possible to distinguish between different watermarks embedded with our method to differentiate between generative models.
  arXiv  Detail & Related papers  (2025-02-12T09:00:48Z)
• ROBIN: Robust and Invisible Watermarks for Diffusion Models with Adversarial Optimization [15.570148419846175]
  Existing watermarking methods face the challenge of balancing robustness and concealment.<n>This paper introduces a watermark hiding process to actively achieve concealment, thus allowing the embedding of stronger watermarks.<n> Experiments on various diffusion models demonstrate the watermark remains verifiable even under significant image tampering.
  arXiv  Detail & Related papers  (2024-11-06T12:14:23Z)
• Shallow Diffuse: Robust and Invisible Watermarking through Low-Dimensional Subspaces in Diffusion Models [13.800130459253543]
  We introduce Shallow Diffuse, a new watermarking technique that embeds robust and invisible watermarks into diffusion model outputs.<n>Our theoretical and empirical analyses show that Shallow Diffuse greatly enhances the consistency of data generation and the detectability of the watermark.
  arXiv  Detail & Related papers  (2024-10-28T14:51:04Z)
• Inevitable Trade-off between Watermark Strength and Speculative Sampling Efficiency for Language Models [63.450843788680196]
  We show that it is impossible to simultaneously maintain the highest watermark strength and the highest sampling efficiency. We propose two methods that maintain either the sampling efficiency or the watermark strength, but not both. Our work provides a rigorous theoretical foundation for understanding the inherent trade-off between watermark strength and sampling efficiency.
  arXiv  Detail & Related papers  (2024-10-27T12:00:19Z)
• Unbiased Watermark for Large Language Models [67.43415395591221]
  This study examines how significantly watermarks impact the quality of model-generated outputs. It is possible to integrate watermarks without affecting the output probability distribution. The presence of watermarks does not compromise the performance of the model in downstream tasks.
  arXiv  Detail & Related papers  (2023-09-22T12:46:38Z)
• Watermarking Images in Self-Supervised Latent Spaces [75.99287942537138]
  We revisit watermarking techniques based on pre-trained deep networks, in the light of self-supervised approaches. We present a way to embed both marks and binary messages into their latent spaces, leveraging data augmentation at marking time.
  arXiv  Detail & Related papers  (2021-12-17T15:52:46Z)
• Fine-tuning Is Not Enough: A Simple yet Effective Watermark Removal Attack for DNN Models [72.9364216776529]
  We propose a novel watermark removal attack from a different perspective. We design a simple yet powerful transformation algorithm by combining imperceptible pattern embedding and spatial-level transformations. Our attack can bypass state-of-the-art watermarking solutions with very high success rates.
  arXiv  Detail & Related papers  (2020-09-18T09:14:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.