Deep Transductive Nonnegative Matrix Factorization for Speech Separation

Liu, Yalin; Guan, Naiyang; Liu, Jie

doi:10.1109/icmla.2017.0-151

Cited by 3 publications

(5 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nonnegative matrix factorization (NMF) preserves the non-negativity property of the magnitude spectrogram of the speech signal [15]. Given a sequence of time-domain signal v(t), its spectrogram is expressed by formula ( 1…”

Section: Nmf For Single-channel Speech Separationmentioning

confidence: 99%

“…Since based on a family of statistical models [10], the NMF is easy to extend in many ways, such as adding sparsity [11], dynamics across time [12][13], and phase awareness [14]. However, these extensions are still limited to monolayer structures, which are difficult to extract non-linear features in speech [15]. According to the big success of deep neural networks in speech recognition, the hierarchical features can represent speech signals well [16].…”

Section: Introductionmentioning

confidence: 99%

“…Based on the multi-layer NMF model, the researchers carried out a further extension. Naiyang Guan proposed a deep transductive NMF model (DTNMF), which incorporated a multilayer structure into NMF and learned a shared dictionary on the source signal of each speaker and the mixture signal to be separated [15]. Chung-Chien Hsu developed the DLNMF by extending the generative bases in LNMF to the discriminative bases which were estimated according to a discriminative criterion [19].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Layered Convolutive Nonnegative Matrix Factorization for Speech Separation

Wang

Huang

et al. 2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Nonnegative matrix factorization (NMF) has attracted significant attention for its good performance in single-channel speech separation. The improved algorithms of NMF have become research hotspots. Layered NMF (LNMF), an improved algorithm, can express the source signal more accurately for its multilayer structure. However, LNMF sometimes performs poorly because it ignores the short-term correlation of speech signals. Based on LNMF and the advantages of Convolutive NMF (CNMF), we proposed a Layered Convolutive NMF(LCNMF) algorithm for single-channel speech separation. The LCNMF corporates the multilayer structure into the NMF and expands the convolution of the top-level NMF model. During the training, NMF is used to learn the non-top-level basis matrices, and CNMF is used to learn the top-level basis matrix, then combined with each single-layer of basis matrix. During the prediction, CNMF is used to separate mixed signals. The results on the dataset MIK-1K showed that LCNMF outperformed NMF and LNMF for separating the mixture of single-channel speech signals. LCNMF improved by 0.019, 1.049dB, 1.305dB, and 0.851dB on average compared with NMF, and improved by 0.007, 0.172dB, 0.090dB, and 0.366dB on average compared with LNMF in sort-term objective intelligibility (STOI), Source to Distortion Ratio (SDR), Source to Interference Ratio (SIR) and Source to Artifacts Ratio (SAR)

show abstract

Section: Nmf For Single-channel Speech Separationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Layered Convolutive Nonnegative Matrix Factorization for Speech Separation

Wang

Huang

et al. 2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

show abstract

“…The algorithm also directly factorizes the basis images of each layer. To obtain the final factorization form in (24), it is necessary to solve the following optimization problem with regularization by hierarchically optimizing the basis images matrix of each layer.…”

Section: The Regularized Deep Nonlinear Non-negative Basis Matrix Facmentioning

confidence: 99%

“…Deep learning based methods have been applied to many practical data analysis recently [21], [22]. Therefore, researchers focus largely on studying the multilayered NMF algorithm in recent years [23], [24], [25], [26], [27], [28]. Ahn et al proposed a multiple NMF (MNMF) network structure in which the deep feature of the sample under the basis images is obtained by continuously performing non-negative factorization on the coefficient matrix [29].…”

Section: Introductionmentioning

confidence: 99%

Deep Non-Negative Matrix Factorization Architecture Based on Underlying Basis Images Learning

Zhao

Wang

Pei

2021

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

The non-negative matrix factorization (NMF) algorithm represents the original image as a linear combination of a set of basis images. This image representation method is in line with the idea of "parts constitute a whole" in human thinking. The existing deep NMF performs deep factorization on the coefficient matrix. In these methods, the basis images used to represent the original image is essentially obtained by factorizing the original images once. To extract features reflecting the deep localization characteristics of images, a novel deep NMF architecture based on underlying basis images learning is proposed for the first time. The architecture learns the underlying basis images by deep factorization on the basis images matrix. The deep factorization architecture proposed in this paper has strong interpretability. To implement this architecture, this paper proposes a deep non-negative basis matrix factorization algorithm to obtain the underlying basis images. Then, the objective function is established with an added regularization term, which directly constrains the basis images matrix to obtain the basis images with good local characteristics, and a regularized deep non-negative basis matrix factorization algorithm is proposed. The regularized deep nonlinear non-negative basis matrix factorization algorithm is also proposed to handle pattern recognition tasks with complex data. This paper also theoretically proves the convergence of the algorithm. Finally, the experimental results show that the deep NMF architecture based on the underlying basis images learning proposed in this paper can obtain better recognition performance than the other state-of-the-art methods.

show abstract