Hidden in Plain Tokens: Simply Robust, Gradient-Free Watermark for Synthetic Audio

• URL: http://arxiv.org/abs/2605.25967v1
• Date: Mon, 25 May 2026 15:43:20 GMT
• Title: Hidden in Plain Tokens: Simply Robust, Gradient-Free Watermark for Synthetic Audio
• Abstract summary: Inference-time watermarks for autoregressive models are unfit for continuous modalities due to discretization inconsistencies.<n>We propose an elegant solution for powerful and robust watermarking of synthetic audio.
• Score: 58.612001688217056
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: As policy catches up with the capabilities of generative AI, watermarking is central to content provenance efforts. Inference-time watermarks for autoregressive models are unfit for continuous modalities due to discretization inconsistencies. Existing methods overcome this by finetuning the modality tokenizers, nullifying the watermark's training-free advantage. In this work, motivated by the vocabulary redundancy of discretization, we propose an elegant solution for powerful and robust watermarking of synthetic audio. We theoretically analyze the impact of token errors on watermark detection, and effectively mitigate them using a reduced vocabulary obtained via community detection. Thorough experiments showcase that our gradient-free method can boost detectability by several orders of magnitude, while also achieving built-in robustness to audio modifications. Broadly, we discover a new state-of-the-art for token-level watermarks in multimedia, which simply arises from the nature of discrete representation learning.

Related papers

• Feature-Aligned Speech Watermarking for Robustness to Reconstruction Distortions [12.934502081353147]
  We propose a feature-aligned watermarking method that aligns the watermark with the original speech feature distribution.<n>We use a pretrained speech to generate a pseudo-speech watermark and fuse it into the spectrogram of the input audio, with VAD loss and perceptual losses guiding embedding.<n>Experiments show that our method maintains imperceptibility comparable to existing approaches while substantially improving robustness under both seen and unseen speech reconstruction models.
  arXiv  Detail & Related papers  (2026-06-10T09:06:37Z)
• More Haste, Less Speed: Weaker Single-Layer Watermark Improves Distortion-Free Watermark Ensembles [58.941305935872265]
  We show that strong watermarks significantly reduce the entropy of the token distribution.<n>We propose a framework that utilizes weaker single-layer watermarks to preserve the entropy required for effective multi-layer ensembling.
  arXiv  Detail & Related papers  (2026-02-12T10:18:16Z)
• Diffusion-Based Image Editing: An Unforeseen Adversary to Robust Invisible Watermarks [4.138397555991069]
  Powerful diffusion-based image generation and editing models can inadvertently remove or distort embedded watermarks.<n>We present a theoretical and empirical analysis demonstrating that diffusion-based image editing can effectively break state-of-the-art robust watermarks.<n>We propose a diffusion-driven attack that uses generative image regeneration to erase watermarks from a given image.
  arXiv  Detail & Related papers  (2025-11-05T16:20:29Z)
• Character-Level Perturbations Disrupt LLM Watermarks [64.60090923837701]
  We formalize the system model for Large Language Model (LLM) watermarking.<n>We characterize two realistic threat models constrained on limited access to the watermark detector.<n>We demonstrate character-level perturbations are significantly more effective for watermark removal under the most restrictive threat model.<n> Experiments confirm the superiority of character-level perturbations and the effectiveness of the Genetic Algorithm (GA) in removing watermarks under realistic constraints.
  arXiv  Detail & Related papers  (2025-09-11T02:50:07Z)
• Robust Watermarks Leak: Channel-Aware Feature Extraction Enables Adversarial Watermark Manipulation [21.41643665626451]
  We propose an attack framework that extracts leakage of watermark patterns using a pre-trained vision model.<n>Unlike prior works requiring massive data or detector access, our method achieves both forgery and detection evasion with a single watermarked image.<n>Our work exposes the robustness-stealthiness paradox: current "robust" watermarks sacrifice security for distortion resistance, providing insights for future watermark design.
  arXiv  Detail & Related papers  (2025-02-10T12:55:08Z)
• Speech Watermarking with Discrete Intermediate Representations [45.892635912641836]
  We propose a novel speech watermarking framework that injects watermarks into the discrete intermediate representations of speech.<n>DiscreteWM achieves state-of-the-art performance in robustness and imperceptibility, simultaneously.<n>Our flexible frame-wise approach can serve as an efficient solution for both voice cloning detection and information hiding.
  arXiv  Detail & Related papers  (2024-12-18T14:57:06Z)
• Image Watermarks are Removable Using Controllable Regeneration from Clean Noise [26.09012436917272]
  A critical attribute of watermark techniques is their robustness against various manipulations.<n>We introduce a watermark removal approach capable of effectively nullifying state-of-the-art watermarking techniques.<n>Our methods offer superior visual consistency/quality and enhanced watermark removal performance compared to existing regeneration approaches.
  arXiv  Detail & Related papers  (2024-10-07T20:04:29Z)
• Duwak: Dual Watermarks in Large Language Models [49.00264962860555]
  We propose, Duwak, to enhance the efficiency and quality of watermarking by embedding dual secret patterns in both token probability distribution and sampling schemes. We evaluate Duwak extensively on Llama2, against four state-of-the-art watermarking techniques and combinations of them.
  arXiv  Detail & Related papers  (2024-03-12T16:25:38Z)
• WatME: Towards Lossless Watermarking Through Lexical Redundancy [58.61972059246715]
  This study assesses the impact of watermarking on different capabilities of large language models (LLMs) from a cognitive science lens. We introduce Watermarking with Mutual Exclusion (WatME) to seamlessly integrate watermarks.
  arXiv  Detail & Related papers  (2023-11-16T11:58:31Z)
• A Resilient and Accessible Distribution-Preserving Watermark for Large Language Models [65.40460716619772]
  Our research focuses on the importance of a textbfDistribution-textbfPreserving (DiP) watermark. Contrary to the current strategies, our proposed DiPmark simultaneously preserves the original token distribution during watermarking. It is detectable without access to the language model API and prompts (accessible), and is provably robust to moderate changes of tokens.
  arXiv  Detail & Related papers  (2023-10-11T17:57:35Z)
• Tree-Ring Watermarks: Fingerprints for Diffusion Images that are Invisible and Robust [55.91987293510401]
  Watermarking the outputs of generative models is a crucial technique for tracing copyright and preventing potential harm from AI-generated content. We introduce a novel technique called Tree-Ring Watermarking that robustly fingerprints diffusion model outputs. Our watermark is semantically hidden in the image space and is far more robust than watermarking alternatives that are currently deployed.
  arXiv  Detail & Related papers  (2023-05-31T17:00:31Z)

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.