Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

Joint Separation and Dereverberation of Reverberant Mixtures with Determined Multichannel Non-negative Matrix Factorization

Abstract: 

This paper proposes an extension of multichannel non-negative matrix factorization (MNMF) that simultaneously solves source separation and dereverberation. While MNMF was originally formulated under an underdetermined problem setting where sources outnumber microphones, a determined counterpart of MNMF, which we call the determined MNMF (DMNMF), has recently been proposed with notable success. This approach is particularly notable in that the optimization process can be more than 30 times faster than the underdetermined version owing to the fact that it involves no matrix inversion computations. However, one drawback as regards all methods based on instantaneous mixture models, including MNMF, is that they are weak against long reverberation. To overcome this drawback, this paper proposes an extension of DMNMF using a frequency-domain convolutive mixture model. The optimization process of the proposed method consists of iteratively updating (i) the spectral parameters of each source using the majorization-minimization algorithm, (ii) the separation matrix using iterative projection, and (iii) the dereverberation filters using multichannel linear prediction. Experimental results showed that the proposed method yielded higher separation performance and dereverberation performance than the baseline method under highly reverberant environments.

up
0 users have voted:

Paper Details

Authors:
Hideaki Kagami, Hirokazu Kameoka, Masahiro Yukawa
Submitted On:
23 April 2018 - 5:00pm
Short Link:
Type:
Presentation Slides
Event:
Presenter's Name:
Hirokazu Kameoka
Paper Code:
4267
Document Year:
2018
Cite

Document Files

Kagami2018ICASSP03.pdf

(35 downloads)

Subscribe

[1] Hideaki Kagami, Hirokazu Kameoka, Masahiro Yukawa, "Joint Separation and Dereverberation of Reverberant Mixtures with Determined Multichannel Non-negative Matrix Factorization", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/3153. Accessed: Jun. 21, 2018.
@article{3153-18,
url = {http://sigport.org/3153},
author = {Hideaki Kagami; Hirokazu Kameoka; Masahiro Yukawa },
publisher = {IEEE SigPort},
title = {Joint Separation and Dereverberation of Reverberant Mixtures with Determined Multichannel Non-negative Matrix Factorization},
year = {2018} }
TY - EJOUR
T1 - Joint Separation and Dereverberation of Reverberant Mixtures with Determined Multichannel Non-negative Matrix Factorization
AU - Hideaki Kagami; Hirokazu Kameoka; Masahiro Yukawa
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/3153
ER -
Hideaki Kagami, Hirokazu Kameoka, Masahiro Yukawa. (2018). Joint Separation and Dereverberation of Reverberant Mixtures with Determined Multichannel Non-negative Matrix Factorization. IEEE SigPort. http://sigport.org/3153
Hideaki Kagami, Hirokazu Kameoka, Masahiro Yukawa, 2018. Joint Separation and Dereverberation of Reverberant Mixtures with Determined Multichannel Non-negative Matrix Factorization. Available at: http://sigport.org/3153.
Hideaki Kagami, Hirokazu Kameoka, Masahiro Yukawa. (2018). "Joint Separation and Dereverberation of Reverberant Mixtures with Determined Multichannel Non-negative Matrix Factorization." Web.
1. Hideaki Kagami, Hirokazu Kameoka, Masahiro Yukawa. Joint Separation and Dereverberation of Reverberant Mixtures with Determined Multichannel Non-negative Matrix Factorization [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/3153