arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2510.01130eess.AS
收藏

用于单耳语音增强的时间图频率表示学习

Learning Time-Graph Frequency Representation for Monaural Speech Enhancement

Tingting Wang, Tianrui Wang, Meng Ge, Qiquan Zhang, Xi Shao

AI总结:

针对现有基于GFT的语音增强方法图拓扑固定、矩阵求逆易产生数值误差与不稳定的问题,提出可学习GFT-SVD框架,用图移位算子和Conv-1D构建可学习图拓扑与傅里叶基,无需矩阵求逆。

摘要:

图傅里叶变换(GFT)近期在语音增强领域展现出良好效果。然而,现有基于GFT的语音增强方法通常采用固定图拓扑构建图傅里叶基,其表示缺乏自适应性与灵活性。此外,基于奇异值分解的GFT(GFT-SVD)和基于特征向量分解的GFT(GFT-EVD)均存在矩阵求逆带来的数值误差与不稳定性问题。受这些局限性启发,本文提出一种简单且有效的可学习GFT-SVD语音增强框架。具体而言,我们利用图移位算子构建可学习图拓扑,并通过一维卷积(Conv-1D)神经层的奇异值矩阵定义可学习图傅里叶基。这一设计无需矩阵求逆,从而避免了相关的数值误差与稳定性问题。

英文摘要:

The Graph Fourier Transform (GFT) has recently demonstrated promising results in speech enhancement. However, existing GFT-based speech enhancement approaches often employ fixed graph topologies to build the graph Fourier basis, whose the representation lacks the adaptively and flexibility. In addition, they suffer from the numerical errors and instability introduced by matrix inversion in GFT based on both Singular Value Decomposition (GFT-SVD) and Eigen Vector Decomposition (GFT-EVD). Motivated by these limitations, this paper propose a simple yet effective learnable GFT-SVD framework for speech enhancement. Specifically, we leverage graph shift operators to construct a learnable graph topology and define a learnable graph Fourier basis by the singular value matrices using 1-D convolution (Conv-1D) neural layer. This eliminates the need for matrix inversion, thereby avoiding the associated numerical errors and stability problem.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑