Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Sound Pattern Recognition: Evaluation of Independent Component Analysis Algorithms for Separation of Voices
KTH, School of Engineering Sciences (SCI).
KTH, School of Engineering Sciences (SCI).
2018 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesisAlternative title
Ljudmönsterigenkänning : Utvärdering av oberoende komponentanalysalgoritmer för separation avröster (Swedish)
Abstract [en]

With computers being used for more applications where commands can be spoken it is useful to findalgorithms which can separate voices from each other so that software can turn spoken words intocommands. In this paper our goal is to describe how Independent Component Analysis (ICA) can beused for separation of voices in cases where we have at least the same number of microphones, atdifferent distances from the speakers, as speakers whose voices we wish to separate, the so called``cocktail party problem". This is done by implementing an ICA algorithm on voice recordingscontaining multiple persons and examining the results. The use of both ICA algorithms result in aclear separation of voices, the advantage of fastICA is that the computations take a fraction of thetime needed for the ML-ICA. Both algorithms can also successfully separate voices when recordingsare made by more microphones than speakers. The algorithms were also able to separate some ofthe voices when there were fewer microphones than speakers which was surprising as thealgorithms have no theoretical guarantee for this.

Abstract [sv]

I detta arbete har vi undersökt hur oberoende komponentanalys algoritmer (ICA) kan användas förseparation av röster där vi har ett varierande antal röster och mikrofoner utplacerade på olikapositioner i ett rum, mer känt som ”cocktailparty problemet”. Detta görs genom att resultatet frånICA-algoritmer appliceras på ljudinspelningar där flera personer talar i mun på varandra. Vi testar ICAalgoritmerna Maximum Likelihood-ICA (ML-ICA) och fastICA. Båda algoritmerna ger goda resultat närdet är minst lika många mikrofoner som talare. Fördelen med fastICA mot ML-ICA är att körtiden ärmycket kortare. Överraskande resultat från båda algoritmerna är att de klarade att separera ut minsten av rösterna när det var fler talare än mikrofoner då detta inte var ett förväntat resultat.

Place, publisher, year, edition, pages
2018. , p. 12
Series
TRITA-SCI-GRU ; 2018-222
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:kth:diva-230746OAI: oai:DiVA.org:kth-230746DiVA, id: diva2:1219135
Examiners
Available from: 2018-06-15 Created: 2018-06-15 Last updated: 2018-06-15Bibliographically approved

Open Access in DiVA

fulltext(904 kB)17 downloads
File information
File name FULLTEXT01.pdfFile size 904 kBChecksum SHA-512
31a43b2a5b9f81a7c3f78d79cbc6b815634c215a202018ed5033bcc772f1a123f6c747d0510058974cc1ecc765db1474443efe7fb3dcb866adc40bf3499f27d9
Type fulltextMimetype application/pdf

By organisation
School of Engineering Sciences (SCI)
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 17 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 7 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf