降低多通道维纳滤波语音增强方法计算复杂度的新策略  

A NOVEL STRATEGY FOR REDUCING COMPUTATIONAL COMPLEXITY OFMULTICHANNEL WIENER FILTERING SPEECH ENHANCEMENT APPROACH

作　　者：涂井先 冀占江 覃桂茳 Tu Jingxian;Ji Zhanjiang;Qin Guijiang(School of Data Science&Software Engineering,Wuzhou University,Wuzhou 543002,Guangxi,China)

机构地区：[1]梧州学院大数据与软件工程学院,广西梧州543002

出　　处：《计算机应用与软件》2023年第11期149-155,共7页Computer Applications and Software

基　　金：广西自然科学基金项目(2018JJB170034);梧州学院校级科研项目(2017B004)。

摘　　要：提出三种改进的多通道维纳滤波算法。其中,每帧信号相邻多个频点共用一个相同线性滤波向量。跟传统算法比,所提出的三种算法有较低的计算复杂度。实验结果表明,以三种经典的客观评价为标准,所提出的三种算法在一些情形下比传统的多通道维纳滤波算法具有更好的去噪性能。This paper proposes three improved multichannel Wiener filtering algorithms.In the proposed algorithms,adjacent multiple frequency bins at each time frame shared the same complex linear filter.Compared with conventional algorithms,the proposed three algorithms had lower computational complexity.Simulation results show that the proposed three algorithms achieve better performance in terms of three classical objective measures in some cases than several conventional multichannel Wiener filtering algorithms.

关 键 词：多通道语音增强 噪声减少 语音失真 维纳滤波 计算复杂度 

分 类 号：TP302.7[自动化与计算机技术—计算机系统结构]