摘要
在语音信号处理领域,子空间降噪算法和自适应算法是经常被用来处理信号的技术手段。分析和研究VSLMS自适应滤波算法和子空间变换的原理和方法[1-2],提出了一种新的组合方式,实现语音降噪算法的进一步改善。组合算法首先使用归一化的子空间算法对语音信号进行子空间层面分解,然后将分离出来的带噪信号作为自适应滤波算法的输入,利用处理因子的迭代,实现对带噪信号的增强处理,最终实现语音处理的优化。实验表明,组合算法的降噪效果,优于传统的使用单个算法进行降噪的方法。
In the field of speech signal processing,subspace algorithm and adaptive algorithm are often used to signal processing technology. Analysis and research on VSLMS adaptive filter algorithm and the principle and method of subspace transformation[1-2],this paper proposes a new way of combining,realizes voice noise reduction algorithm,which is further improved. Combination algorithm firstly it used the normalized subspace algorithm for speech signal subspace decomposition,and then separated the signal with noise as input of adaptive filter algorithm,using iteration of processing factors,realizes the signal with noise enhancement processing,optimization of speech processing. The experiments show that the method effects some extent,is superior to the traditional method of using a single algorithm for noise reduction.
出处
《信息技术》
2015年第10期201-205,共5页
Information Technology
