Minimum-Volume Multichannel Nonnegative Matrix Factorization for Blind Audio Source Separation

Wang, Jianyu; Guan, Shanzheng; Liu, Shupei; Zhang, Xiaolei

doi:10.1109/taslp.2021.3120603

Cited by 7 publications

(7 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…4. ONMF [36]: ONMF is based on the orthogonal non-negative matrix factorization model (orthogonal NMF, ONMF). The basic idea is to impose an orthogonal constraint on W based on the NMF model A-W H 2 F so that W T W = I.…”

Section: Comparison Methodsmentioning

confidence: 99%

Graph Convolutional Networks based Non-Negative Matrix Factorization aware Community detection

Scholar

Professor

2023

Preprint

View full text Add to dashboard Cite

Nonnegative matrix factorization (NMF) is widely used in community discovery because of its effectiveness and easy interpretability. However, most of the existing NMF-based community detection methods are linear. They cannot effectively deal with the nonlinear characteristics of complex networks, resulting in further improvement in community detection performance. Aiming at this problem, a convolution graph network (GCN) enhanced nonlinear NMF community discovery method NMFGCN is proposed. NMFGCN consists of two main modules: GCN and NMF, where GCN is used to learn network node representations, and NMF uses node representations as input to obtain network community representations. In addition, a joint optimization method is proposed to train NMFGCN, which enables NMFGCN to have nonlinear feature representation capabilities and enables GCN and NMF to promote each other and obtain better community segmentation results. Many experiments on artificial synthetic networks and entire networks show that NMFGCN is superior to current NMF-based community discovery methods , thus proving that NMFGCN can improve the performance of NMF community discovery methods. In addition, NMFGCN also outperforms Deep Walk and LINE standard graph representation learning methods.

show abstract

Section: Comparison Methodsmentioning

confidence: 99%

Graph Convolutional Networks based Non-Negative Matrix Factorization aware Community detection

Scholar

Professor

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Since NMF is capable of interpreting the local properties of the image, NMF is proposed as face recognition technology initially [28]. Recently, NMF has been studied for blind source separation [29,30] and NR, on account of sound that can be converted to the spectrogram form.…”

Section: The Nmf-based Nr Techniquementioning

confidence: 99%

Adaptive Noise Reduction Algorithm Based on SPP and NMF for Environmental Sound Event Recognition under Low-SNR Conditions

Luo

Guo

Wang

et al. 2023

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

Nonnegative matrix factorization- (NMF-) based noise reduction methods can effectively improve the performance of environmental sound recognition. However, when the environmental sound overlaps highly with the noise, the spectral line loss and noise residue will occur in the low signal-to-noise ratio (SNR) condition. An adaptive noise reduction algorithm was proposed in this paper. First, noisy environmental sound is separated into estimated noise and environmental sound using NMF. Then, the estimated noise is used to calculate sound presence probability (SPP), which is adapted to decrease spectral line loss and achieve accurately estimated noise. Subsequently, the estimated noise combines with noisy environmental sound to obtain the estimated environmental sound. Finally, SPP is applied to reduce residual noise in the estimated environmental sound and reconstruct the environmental sound. The simulation results demonstrate that the proposed algorithm outperforms the traditional algorithms and NMF-based methods in terms of perceptual evaluation of speech quality (PESQ) and global SNR with increase of X% and X%, respectively. Moreover, the proposed method can effectively improve the environmental sound recognition rate. Particularly, the proposed method makes a 16.2% increase of F1-score in car horn recognition under the realistic acoustic condition.

show abstract

“…It is widely known that the priori information of the source model plays an important role in improving MBSS performance [28,29,19]. While they have been widely studied, the MNMF model given in ( 4) and the ILRMA model given in (6) did not consider the sparse structure of the source prior information in the source model.…”

Section: Cost Functionmentioning

confidence: 99%

“…All the coefficients of ρ n in s-MNMF and s-ILRMA are set to 10 and all the coefficients of µ n in s-MNMF and s-ILRMA are set to 0.05. Besides s-MNMF and s-ILRMA, the following algorithms are also evaluated for the purpose of comparison: s-ILRMA with AuxIVA [42], MNMF [3], m-MNMF [19], ILRMA [6], tILRMA [12], sub-Gaussian distributed ILRMA (sGD-ILRMA) [14] and m-ILRMA [19]. Signal-to-distortion ratio (SDR) and source-to-interference ratio (SIR) [43] are adopted as the performance metrics.…”

Section: Simulation Setupmentioning

confidence: 99%

“…To avoid parameter initialization sensitivity, the so-called t-ILRMA [12], GGD-ILRMA [13,14,15,16], and t-MNMF [17] were developed, which generalize the distribution of source model. To further improve the separation performance, probabilistic sparse distribution was introduced in [18,19] to model the dictionary and activation matrices of source spectrograms though this model is still insufficient for accurately modeling the sources in MBSS [8]. Several methods were then developed to leverage the sparseness of the sources in time-frequency domain [20,21,22,23], which have demonstrated some potential to enhance separation performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Effects of Shear Layer Characteristics on Acoustic Propagation and Source Localization

et al. 2019

View full text Add to dashboard Cite

The shear layer characteristics of an open-jet acoustic wind tunnel are of key importance on measurements of aeroacoustics. The effects of thickness, spreading angle and strength of shear layer on acoustic propagation and source localization are investigated through the mean/spreading shear layer with a self-similar velocity distribution. Based on the shear flow, the acoustic propagation is computed by the linearized Euler equations via a source term, and then source localization is obtained from beamforming technique combined with the theory of Amiet. Results show that the numerical method can precisely capture the refraction and reflection after sound traversing shear layer. The thickness, spreading angle and strength of the shear layer exerts little effects on the refracted region where sound wave nearly vertical incident, while mainly influence the corresponding up/downstream region in terms of phase change. Increment of thickness, spreading angle and strength of the shear layer increases the acoustic difference between the shear layer with and without thickness, and produces a larger error of source localization downstream of the actual position.

show abstract

Minimum-Volume Multichannel Nonnegative Matrix Factorization for Blind Audio Source Separation

Cited by 7 publications

References 47 publications

Graph Convolutional Networks based Non-Negative Matrix Factorization aware Community detection

Graph Convolutional Networks based Non-Negative Matrix Factorization aware Community detection

Adaptive Noise Reduction Algorithm Based on SPP and NMF for Environmental Sound Event Recognition under Low-SNR Conditions

Effects of Shear Layer Characteristics on Acoustic Propagation and Source Localization

Contact Info

Product

Resources

About