Time-domain blind audio source separation using advanced component clustering and reconstruction

被引：17
|
作者：
Koldovsky, Zbynek [1 ]
Tichavsky, Petr [2 ]
机构：
[1] Tech Univ Liberec, Studentska 2, Liberec 46117, Czech Republic
[2] Inst Informat Theory & Automat, Prague, Czech Republic
来源：
2008 HANDS-FREE SPEECH COMMUNICATION AND MICROPHONE ARRAYS | 2008年
关键词：
D O I：
10.1109/HSCMA.2008.4538725
中图分类号：
O42 [声学];
学科分类号：
070206 ; 082403 ;
摘要：
We present a novel time-domain method for blind separation of convolutive mixture of audio sources (the cocktail party problem). The method allows efficient separation with good signal-to-interference ratio (SIR) and signal-to-distortion ratio (SDR) using short data segments only. In practice, we are able to separate 2-4 speakers from audio recording of the length less than 6000 samples, which is less than 1 s in the 8 kHz sampling. The average time needed to process the data with filter of the length 20 was 2.2 seconds in Matlab v. 7.2 on an ordinary PC with 3GHz processor.
引用
收藏
页码：216 / +
页数：2
相关论文
共 50 条
  • [1] Time-Domain Blind Audio Source Separation using Advanced ICA Methods
    Koldovsky, Zbynek
    Tichavsky, Petr
    [J]. INTERSPEECH 2007: 8TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION, VOLS 1-4, 2007, : 1861 - +
  • [2] Fuzzy Clustering of Independent Components within Time-Domain Blind Audio Source Separation Method
    Malek, Jiri
    Koldovsky, Zbynek
    [J]. 2011 10TH INTERNATIONAL WORKSHOP ON ELECTRONICS, CONTROL, MEASUREMENT AND SIGNALS (ECMS), 2011, : 44 - 49
  • [3] A Time-Domain Blind Source Separation by the Newton Method
    Matsuoka, Kiyotoshi
    Itahashi, Takashi
    [J]. ISCAS: 2009 IEEE INTERNATIONAL SYMPOSIUM ON CIRCUITS AND SYSTEMS, VOLS 1-5, 2009, : 2790 - 2793
  • [4] A method of underdetermined blind source separation in time-domain
    Wang Rongjie
    Zhan Yiju
    Zhou Haifeng
    [J]. INTERNATIONAL JOURNAL OF ELECTRONICS, 2012, 99 (04) : 543 - 555
  • [5] Subband Blind Audio Source Separation Using a Time-Domain Algorithm and Tree-Structured QMF Filter Bank
    Koldovsky, Zbynek
    Tichavsky, Petr
    Malek, Jiri
    [J]. LATENT VARIABLE ANALYSIS AND SIGNAL SEPARATION, 2010, 6365 : 25 - +
  • [6] SPARSE GAUSSIAN PROCESS AUDIO SOURCE SEPARATION USING SPECTRUM PRIORS IN THE TIME-DOMAIN
    Alvarado, Pablo A.
    Alvarez, Mauricio A.
    Stowell, Dan
    [J]. 2019 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 2019, : 995 - 999
  • [7] Time-Domain Blind Audio Source Separation Method Producing Separating Filters of Generalized Feedforward Structure
    Koldovsky, Zbynek
    Tichavsky, Petr
    Malek, Jiri
    [J]. LATENT VARIABLE ANALYSIS AND SIGNAL SEPARATION, 2010, 6365 : 17 - +
  • [8] Time-domain detection with blind separation of unknown source signals
    Li, Guan Nan
    Ma, Wei
    Zhang, Ning
    Zhu, Liang
    [J]. ISAPE 2008: THE 8TH INTERNATIONAL SYMPOSIUM ON ANTENNAS, PROPAGATION AND EM THEORY, PROCEEDINGS, VOLS 1-3, 2008, : 1526 - +
  • [9] On the causality problem in time-domain blind source separation and deconvolution algorithms
    Aichner, R
    Buchner, H
    Kellermann, W
    [J]. 2005 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, VOLS 1-5: SPEECH PROCESSING, 2005, : 181 - 184
  • [10] Time-Domain Audio Source Separation With Neural Networks Based on Multiresolution Analysis
    Nakamura, Tomohiko
    Kozuka, Shihori
    Saruwatari, Hiroshi
    [J]. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING, 2021, 29 : 1687 - 1701
12345