Structure from sound with incomplete data

Miranda Krekovic, Gilles Baechler, Ivan Dokmanic, Martin Vetterli

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, we consider the problem of jointly localizing a microphone array and identifying the direction of arrival of acoustic events. Under the assumption that the sources are in the far field, this problem can be formulated as a constrained low-rank matrix factorization with an unknown column offset. Our focus is on handling missing entries, particularly when the measurement matrix does not contain a single complete column. This case has not received attention in the literature and is not handled by existing algorithms, however it is prevalent in practice. We propose an iterative algorithm that works with pairwise differences between the measurements eliminating the dependence on the unknown offset. We demonstrate state-of-the-art performance both in terms of accuracy and versatility.

Original languageEnglish (US)
Title of host publication2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages3539-3543
Number of pages5
ISBN (Print)9781538646588
DOIs
StatePublished - Sep 10 2018
Event2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Calgary, Canada
Duration: Apr 15 2018Apr 20 2018

Publication series

NameICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume2018-April
ISSN (Print)1520-6149

Other

Other2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018
Country/TerritoryCanada
CityCalgary
Period4/15/184/20/18

Keywords

  • Far field
  • Low-rank matrix factorization
  • Missing and uncertain data
  • Node localization
  • Sensor array self-calibration

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Structure from sound with incomplete data'. Together they form a unique fingerprint.

Cite this