Title: SyncGuard: Robust Audio Watermarking Capable of Countering Desynchronization Attacks Show Chinese title

Title: SyncGuard：能够抵御不同步攻击的鲁棒音频水印技术

Abstract: Audio watermarking has been widely applied in copyright protection and source tracing. However, due to the inherent characteristics of audio signals, watermark localization and resistance to desynchronization attacks remain significant challenges. In this paper, we propose a learning-based scheme named SyncGuard to address these challenges. Specifically, we design a frame-wise broadcast embedding strategy to embed the watermark in arbitrary-length audio, enhancing time-independence and eliminating the need for localization during watermark extraction. To further enhance robustness, we introduce a meticulously designed distortion layer. Additionally, we employ dilated residual blocks in conjunction with dilated gated blocks to effectively capture multi-resolution time-frequency features. Extensive experimental results show that SyncGuard efficiently handles variable-length audio segments, outperforms state-of-the-art methods in robustness against various attacks, and delivers superior auditory quality. Show Chinese abstract

Abstract: 音频水印技术已被广泛应用于版权保护和源追踪。 然而，由于音频信号的固有特性，水印定位和抗不同步攻击仍是一个重大挑战。 在本文中，我们提出了一种名为SyncGuard的学习方案来解决这些挑战。 具体而言，我们设计了一种逐帧广播嵌入策略，以在任意长度的音频中嵌入水印，增强时间独立性，并在水印提取过程中消除定位需求。 为了进一步提高鲁棒性，我们引入了一个精心设计的失真层。 此外，我们结合扩张残差块和扩张门控块，以有效捕捉多分辨率时频特征。 大量的实验结果表明，SyncGuard能够高效处理变长音频段，在对抗各种攻击的鲁棒性方面优于最先进的方法，并且提供了优越的听觉质量。