XAttnMark: Learning Robust Audio Watermarking with Cross-Attention

TL;DR

XAttnMark is a novel audio watermarking method that uses cross-attention and psychoacoustic masking to achieve robust detection and accurate attribution in generative and edited audio.

Contribution

It introduces a cross-attention mechanism with parameter sharing and a psychoacoustic masking loss for improved robustness and imperceptibility in audio watermarking.

Findings

Achieves state-of-the-art detection and attribution performance.

Demonstrates robustness against various audio transformations.

Enhances imperceptibility with psychoacoustic masking loss.

Abstract

The rapid proliferation of generative audio synthesis and editing technologies has raised significant concerns about copyright infringement, data provenance, and the spread of misinformation through deepfake audio. Watermarking offers a proactive solution by embedding imperceptible, identifiable, and traceable marks into audio content. While recent neural network-based watermarking methods like WavMark and AudioSeal have improved robustness and quality, they struggle to achieve both robust detection and accurate attribution simultaneously. This paper introduces Cross-Attention Robust Audio Watermark (XAttnMark), which bridges this gap by leveraging partial parameter sharing between the generator and the detector, a cross-attention mechanism for efficient message retrieval, and a temporal conditioning module for improved message distribution. Additionally, we propose a psychoacoustic-aligned temporal-frequency masking loss that captures fine-grained auditory masking effects, enhancing watermark imperceptibility. Our approach achieves state-of-the-art performance in both detection and attribution, demonstrating superior robustness against a wide range of audio transformations, including challenging generative editing with strong editing strength. The project webpage is available at https://liuyixin-louis.github.io/xattnmark/.