摘要
传统广义旁瓣抵消器在阻塞目标信号时,目标信号对消及波束形成方向固定,缺乏灵活性,为此,本文提出基于麦克风阵列技术的一种新型广义旁瓣抵消器自适应语音增强方法。该方法利用自适应阻塞矩阵级联自适应抵消器来估计出维纳滤波器输出信号中的残余噪声,并与输出信号进行对消,然后后续一个谱减法进一步去除可能留有的残余噪声。分析了改进算法的性能,并与传统广义旁瓣抵消器进行了仿真比较。
When the traditional generalized sidelobe canceller（GSC） blockes target signal, the target signal cancellation and the beam-forming direction of the conventional is fixed and lack of flexibility problem, this paper presents a novel generalized sidelobe canceller（GSC） adaptive speech enhancement method. In this method, an adaptive blocking matrix cascaded an adaptive canceller is used to estimates the residual noise of the output signal in Wiener filters,and cancellation is conducted with the output signal. Then a spectral subtraction is followed to further remove residual noise. Compare with the simulation of the GSC algorithm and traditional methods.
作者
赵益波
徐进
阮玮琪
蒋祎
Zhao Yibo Xu Jin Ruan Weiqi Jiang Yi(School of Electronic ＆ Information Engineering,Nanjing University of Information Science ＆ Technology,Nanjing 210044, China Jiangsu Collaborative Innovation Center on Atmospheric Environment and Equipment Nanjing 210044 China)
出处
《科技通报》
北大核心
2016年第9期172-176,共5页
Bulletin of Science and Technology
基金
江苏省高校优势学科建设工程资助项目
江苏省高校自然科学基金(14KJB120007
13KJA510001)
江苏省重点大学生创新项目(201410300023)
江苏省高校科研成果产业化推进项目(JHB2012-9)
优秀博士论文作者专项资金(200753)
国家自然科学基金(11504176)
