A time domain algorithm for blind separation of convolutive sound mixtures and L1 constrainted minimization of cross correlations

Jie Liu Jack Xin Yingyong Qi Fan-Gang Zheng

Geometric Modeling and Processing mathscidoc:1912.43861

Communications in Mathematical Sciences, 7, (1), 109-128, 2009
A time domain blind source separation algorithm of convolutive sound mixtures is studied based on a compact partial inversion formula in closed form. An L1-constrained minimization problem is formulated to find demixing filter coefficients for source separation while capturing scaling invariance and sparseness of solutions. The minimization aims to reduce (lagged) cross correlations of the mixture signals, which are modeled stochastically. The problem is non-convex, however it is put in a nonlinear least squares form where the robust and convergent Levenberg-Marquardt iterative method is applicable to compute local minimizers. Efficiency is achieved in recovering lower dimensional demixing filter solutions than the physical ones. Computations on recorded and synthetic mixtures show satisfactory performance, and are compared with other iterative methods.
No keywords uploaded!
[ Download ] [ 2019-12-24 21:02:55 uploaded by Jack_Xin ] [ 451 downloads ] [ 0 comments ]
@inproceedings{jie2009a,
  title={A time domain algorithm for blind separation of convolutive sound mixtures and L1 constrainted minimization of cross correlations},
  author={Jie Liu, Jack Xin, Yingyong Qi, and Fan-Gang Zheng},
  url={http://archive.ymsc.tsinghua.edu.cn/pacm_paperurl/20191224210255064837425},
  booktitle={Communications in Mathematical Sciences},
  volume={7},
  number={1},
  pages={109-128},
  year={2009},
}
Jie Liu, Jack Xin, Yingyong Qi, and Fan-Gang Zheng. A time domain algorithm for blind separation of convolutive sound mixtures and L1 constrainted minimization of cross correlations. 2009. Vol. 7. In Communications in Mathematical Sciences. pp.109-128. http://archive.ymsc.tsinghua.edu.cn/pacm_paperurl/20191224210255064837425.
Please log in for comment!
 
 
Contact us: office-iccm@tsinghua.edu.cn | Copyright Reserved