摘要
针对独立矢量分析(IVA)算法初始分离矩阵取值对分离性能影响较大的局限性,提出了基于回溯搜索优化的卷积混合语音盲分离算法。采用频域各频率点IVA分离信号的复数峭度和作为目标函数,利用回溯搜索优化算法(BSA)对初始分离矩阵进行优化调整,更好地实现了语音信号的盲分离。在分离过程中,采用复Givens旋转变换原理将对分离矩阵的求解转化为对旋转角度的求解,有效减少了BSA的参数编码维数,降低了优化求解难度。针对语音信号的卷积混合分离实验表明,该算法具有良好的分离效果,其分离性能较之基本IVA算法显著提升。
Aiming to overcome the limitation of initial separation matrix selection in Independent Vector Analysis(IVA),a convolutive blind speech separation algorithm based on backtracking search optimization is proposed.The sum ofcomplex kurtosis of separated signals in each frequency point from IVA is used as the objective function.The BacktrackingSearch Optimization Algorithm(BSA)is used to adjust the initial separation matrix for better separation.In the separationprocess,complex Givens rotation transformation is used to transform separation matrix to rotation angle for reducing thecoding dimension of BSA and the difficulty of optimization decreases.The blind speech separation experiments forconvolutive mixture signals indicate that the proposed algorithm performs excellent separation results and the separationproperty is better than the basic IVA algorithm.
作者
陈雷
韩大伟
郭艳菊
李媛媛
贾志成
CHEN Lei;HAN Dawei;GUO Yanju;LI Yuanyuan;JIA Zhicheng(School of Information Engineering, Tianjin University of Commerce, Tianjin 300134, China;School of Precision Instrument and Opto-Electronics Engineering, Tianjin University, Tianjin 300072, China;School of Electronic Information Engineering, Hebei University of Technology, Tianjin 300401, China)
出处
《计算机工程与应用》
CSCD
北大核心
2017年第15期137-143,共7页
Computer Engineering and Applications
基金
国家自然科学基金(No.61401307)
中国博士后科学基金(No.2014M561184)
天津市应用基础与前沿技术研究计划(No.15JCYBJC17100)
天津市科技特派员项目(No.16JCTPJC48400)
关键词
语音盲分离
回溯搜索优化算法
卷积混合
独立矢量分析
Givens旋转变换
speech blind separation
backtracking search optimization algorithm
convolutive mixture
independent vector analysis
Givens rotation transformation
