A GMM sound source model for blind speech separation in under-determined conditions

Authors:
Yasuharu Hirasawa;Naoki Yasuraoka;Toru Takahashi;Tetsuya Ogata;Hiroshi G. Okuno
Affiliations:
Graduate School of Informatics, Kyoto University, Kyoto, Japan;Graduate School of Informatics, Kyoto University, Kyoto, Japan;Graduate School of Informatics, Kyoto University, Kyoto, Japan;Graduate School of Informatics, Kyoto University, Kyoto, Japan;Graduate School of Informatics, Kyoto University, Kyoto, Japan
Venue:
LVA/ICA'12 Proceedings of the 10th international conference on Latent Variable Analysis and Signal Separation
Year:
2012

Citing 5
Cited 1

Multichannel nonnegative matrix factorization in convolutive mixtures for audio source separation

IEEE Transactions on Audio, Speech, and Language Processing
Under-determined reverberant audio source separation using a full-rank spatial covariance model

IEEE Transactions on Audio, Speech, and Language Processing - Special issue on processing reverberant speech: methodologies and applications
The 2010 signal separation evaluation campaign (SiSEC2010): audio source separation

LVA/ICA'10 Proceedings of the 9th international conference on Latent variable analysis and signal separation
Performance measurement in blind audio source separation

IEEE Transactions on Audio, Speech, and Language Processing
Underdetermined Convolutive Blind Source Separation via Frequency Bin-Wise Clustering and Permutation Alignment

IEEE Transactions on Audio, Speech, and Language Processing

The 2011 signal separation evaluation campaign (SiSEC2011): - audio source separation -

LVA/ICA'12 Proceedings of the 10th international conference on Latent Variable Analysis and Signal Separation

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper focuses on blind speech separation in under-determined conditions, that is, in the case when there are more sound sources than microphones. We introduce a sound source model based on the Gaussian mixture model (GMM) to represent a speech signal in the time-frequency domain, and derive rules for updating the model parameters using the auxiliary function method. Our GMM sound source model consists of two kinds of Gaussians: sharp ones representing harmonic parts and smooth ones representing nonharmonic parts. Experimental results reveal that our method outperforms the method based on non-negative matrix factorization (NMF) by 0.7dB in the signal-to-distortion ratio (SDR), and by 1.7dB in the signal-to-interference ratio (SIR). This means that our method effectively removes interference coming from other talkers.