基于语音参数模型的抗同步攻击音频水印算法

Audio Watermark Algorithm Robust to Desynchronization Based on Speech Parameter Model

同步是音频水印需要解决的关键问题之一,目前的音频水印算法还缺乏简单有效的同步机制。该文提出一种基于语音参数模型的自同步水印算法,在保证不可感知性和稳健性的良好平衡下实现水印的盲检测。仿真实验表明,该同步机制算法复杂度低,稳健性强,实现简单,对实际应用有重要意义。

Synchronization is one of the key issues of audio watermark. Current audio watermark algorithms lack effective techniques for synchronization. Blind detection is realized with the balance between inaudibility and robustness. Experimental results show the synchronization scheme is low-complex, robust and simple for application.