Multiple Speech Source Separation with Non-Sparse Components Recovery by Using Dual Similarity Determination

Maoshen JIA; Jundai SUN; Feng DENG; Junyue SUN

doi:10.1587/transinf.2016IIP0019

Multiple Speech Source Separation with Non-Sparse Components Recovery by Using Dual Similarity Determination

Maoshen JIA, Jundai SUN, Feng DENG, Junyue SUN

Full Text Views

0

Share
Cite this

Summary :

In this work, a multiple source separation method with joint sparse and non-sparse components recovery is proposed by using dual similarity determination. Specifically, a dual similarity coefficient is designed based on normalized cross-correlation and Jaccard coefficients, and its reasonability is validated via a statistical analysis on a quantitative effective measure. Thereafter, by regarding the sparse components as a guide, the non-sparse components are recovered using the dual similarity coefficient. Eventually, a separated signal is obtained by a synthesis of the sparse and non-sparse components. Experimental results demonstrate the separation quality of the proposed method outperforms some existing BSS methods including sparse components separation based methods, independent components analysis based methods and soft threshold based methods.

Publication: IEICE TRANSACTIONS on Information Vol.E101-D No.4 pp.925-932

Publication Date: 2018/04/01

Publicized: 2018/01/19

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2016IIP0019

Type of Manuscript: Special Section PAPER (Special Section on Intelligent Information and Communication Technology and its Applications to Creative Activity Support)

Category: Elemental Technologies for human behavior analysis

Authors

Maoshen JIA
  Beijing University of Technology
Jundai SUN
  Beijing University of Technology
Feng DENG
  Dolby Laboratories
Junyue SUN
  Dialog Semiconductor

Keyword

blind source separation, non-sparse component, similarity coefficient, intelligent systems

Cite this

Copy

Maoshen JIA, Jundai SUN, Feng DENG, Junyue SUN, "Multiple Speech Source Separation with Non-Sparse Components Recovery by Using Dual Similarity Determination" in IEICE TRANSACTIONS on Information, vol. E101-D, no. 4, pp. 925-932, April 2018, doi: 10.1587/transinf.2016IIP0019.
Abstract: In this work, a multiple source separation method with joint sparse and non-sparse components recovery is proposed by using dual similarity determination. Specifically, a dual similarity coefficient is designed based on normalized cross-correlation and Jaccard coefficients, and its reasonability is validated via a statistical analysis on a quantitative effective measure. Thereafter, by regarding the sparse components as a guide, the non-sparse components are recovered using the dual similarity coefficient. Eventually, a separated signal is obtained by a synthesis of the sparse and non-sparse components. Experimental results demonstrate the separation quality of the proposed method outperforms some existing BSS methods including sparse components separation based methods, independent components analysis based methods and soft threshold based methods.
URL: https://globals.ieice.org/en_transactions/information/10.1587/transinf.2016IIP0019/_p

Copy

@ARTICLE{e101-d_4_925,
author={Maoshen JIA, Jundai SUN, Feng DENG, Junyue SUN, },
journal={IEICE TRANSACTIONS on Information},
title={Multiple Speech Source Separation with Non-Sparse Components Recovery by Using Dual Similarity Determination},
year={2018},
volume={E101-D},
number={4},
pages={925-932},
abstract={In this work, a multiple source separation method with joint sparse and non-sparse components recovery is proposed by using dual similarity determination. Specifically, a dual similarity coefficient is designed based on normalized cross-correlation and Jaccard coefficients, and its reasonability is validated via a statistical analysis on a quantitative effective measure. Thereafter, by regarding the sparse components as a guide, the non-sparse components are recovered using the dual similarity coefficient. Eventually, a separated signal is obtained by a synthesis of the sparse and non-sparse components. Experimental results demonstrate the separation quality of the proposed method outperforms some existing BSS methods including sparse components separation based methods, independent components analysis based methods and soft threshold based methods.},
keywords={},
doi={10.1587/transinf.2016IIP0019},
ISSN={1745-1361},
month={April},}

Copy

TY - JOUR
TI - Multiple Speech Source Separation with Non-Sparse Components Recovery by Using Dual Similarity Determination
T2 - IEICE TRANSACTIONS on Information
SP - 925
EP - 932
AU - Maoshen JIA
AU - Jundai SUN
AU - Feng DENG
AU - Junyue SUN
PY - 2018
DO - 10.1587/transinf.2016IIP0019
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E101-D
IS - 4
JA - IEICE TRANSACTIONS on Information
Y1 - April 2018
AB - In this work, a multiple source separation method with joint sparse and non-sparse components recovery is proposed by using dual similarity determination. Specifically, a dual similarity coefficient is designed based on normalized cross-correlation and Jaccard coefficients, and its reasonability is validated via a statistical analysis on a quantitative effective measure. Thereafter, by regarding the sparse components as a guide, the non-sparse components are recovered using the dual similarity coefficient. Eventually, a separated signal is obtained by a synthesis of the sparse and non-sparse components. Experimental results demonstrate the separation quality of the proposed method outperforms some existing BSS methods including sparse components separation based methods, independent components analysis based methods and soft threshold based methods.
ER -