Publication detail

Undertermined Blind Source Separation Using Linear Separation System

SMÉKAL, Z. ČERMÁK, J.

Original Title

Undertermined Blind Source Separation Using Linear Separation System

English Title

Undertermined Blind Source Separation Using Linear Separation System

Type

journal article - other

Language

en

Original Abstract

In automatic speech recognition and speech modal analysis, good quality of input speech signal is often required. The hit rate of recognizers is lowered by degradation of speech quality due to the noise. Blind source separation can be used to enhance the speech signal as a part of preprocessing techniques. This paper presents a multi channel linear blind source separation method that can be applied even in underdetermined case (UDC) i.e. when the number of source signals is higher than the number of sensors. The experiments show that our system outperforms conventional time-frequency binary masking (TFBM) in both determined and underdetermined cases.

English abstract

In automatic speech recognition and speech modal analysis, good quality of input speech signal is often required. The hit rate of recognizers is lowered by degradation of speech quality due to the noise. Blind source separation can be used to enhance the speech signal as a part of preprocessing techniques. This paper presents a multi channel linear blind source separation method that can be applied even in underdetermined case (UDC) i.e. when the number of source signals is higher than the number of sensors. The experiments show that our system outperforms conventional time-frequency binary masking (TFBM) in both determined and underdetermined cases.

Keywords

Array signal processing, beamforming, blind source separation, speech processing, time-frequency binary masking.

RIV year

2009

Released

06.03.2009

Publisher

Springer Verlag

Location

Berlin

Pages from

300

Pages to

305

Pages count

5

BibTex


@article{BUT47335,
  author="Zdeněk {Smékal} and Jan {Čermák}",
  title="Undertermined Blind Source Separation Using Linear Separation System",
  annote="In automatic speech recognition and speech modal analysis, good quality of input speech signal is often required. The hit rate of recognizers is lowered by degradation of speech quality due to the noise. Blind source separation can be used to enhance the speech signal as a part of preprocessing techniques. This paper presents a multi channel linear blind source separation method that can be applied even in underdetermined case (UDC) i.e. when the number of source signals is higher than the number of sensors. The experiments show that our system outperforms conventional time-frequency binary masking (TFBM) in both determined and underdetermined cases.",
  address="Springer Verlag",
  chapter="47335",
  institution="Springer Verlag",
  journal="Lecture Notes in Computer Science (IF 0,513)",
  number="5398",
  volume="2009",
  year="2009",
  month="march",
  pages="300--305",
  publisher="Springer Verlag",
  type="journal article - other"
}