标题:Component separation with gaussian moments
作者:Zhao, Yongjian
通讯作者:Zhao, Yongjian
作者机构:[Zhao, Yongjian ] Shandong University, Weihai; 264209, China
会议名称:International Conference on Applications and Techniques in Cyber Intelligence, ATCI 2018
会议日期:11 July 2018 through 13 July 2018
来源:Advances in Intelligent Systems and Computing
出版年:2019
卷:842
页码:1337-1342
DOI:10.1007/978-3-319-98776-7_166
关键词:Component; Independence; Kurtosis; Moment; Separation; Source
摘要:Blind signal processing (BSS) is a fairly new and generally applicable technique. A very intuitive and important principle for the BSS problem is to maximize/minimize non-Gaussianity. Gaussian moments are introduced here as a quantitative measure of non-Gaussianity for a random variable. A proper contrast function is presented correspondingly that has not asymptotic bias even in the noisy context. After maximization of the contrast function, an improved BSS algorithm is introduced correspondingly. Computer simulations demonstrate the efficiency of the proposed approach. © Springer Nature Switzerland AG 2019.
收录类别:EI;SCOPUS
资源类型:会议论文;期刊论文
原文链接:https://www.scopus.com/inward/record.uri?eid=2-s2.0-85056828172&doi=10.1007%2f978-3-319-98776-7_166&partnerID=40&md5=5087d27781c0d5dbe004f19ae2c45e10
TOP