Multichannel nonnegative matrix factorization for ego-noise suppression

Conference contribution
(Conference Contribution)

Publication Details

Author(s): Haubner T, Schmidt A, Kellermann W
Publication year: 2018
Pages range: 136-140
ISBN: 978-3-8007-4767-2


A mobile robot generates significant self-induced noise,
e.g., by its joints and moving parts of its body, which is
referred to as ego-noise. Ego-noise is of crucial impor-
tance in robot audition since it massively corrupts the mi-
crophone signals and thus degrades the robot’s capability
to interact intuitively with its acoustic environment. There-
fore, appropriate mechanisms for ego-noise suppression
are required. In this paper, we investigate different multi-
channel Nonnegative Matrix Factorization (MNMF) algo-
rithms with respect to their suitability for ego-noise sup-
pression, propose a modified semi-supervised MNMF ap-
proach and evaluate its performance.

FAU Authors / FAU Editors

Haubner, Thomas
Professur für Nachrichtentechnik
Kellermann, Walter Prof. Dr.-Ing.
Professur für Nachrichtentechnik
Schmidt, Alexander
Professur für Nachrichtentechnik

How to cite

Haubner, T., Schmidt, A., & Kellermann, W. (2018). Multichannel nonnegative matrix factorization for ego-noise suppression. (pp. 136-140). Oldenburg, DE.

Haubner, Thomas, Alexander Schmidt, and Walter Kellermann. "Multichannel nonnegative matrix factorization for ego-noise suppression." Proceedings of the ITG Conference on Speech Communication, Oldenburg 2018. 136-140.


Last updated on 2018-16-10 at 12:38