Monaural Sound Source Separation by Nonnegative Matrix Factorization With Temporal Continuity and Sparseness Criteria

By Tuomas Virtanen. IEEE Transactions on Audio, Speech, and Language Processing, 2007

1,106 citations recorded by the source corpus.

Subfields: Computational Mechanics, Signal Processing

View Paper: https://www.semanticscholar.org/paper/fe27cac60301c52397c1ce150abd7706afa007cd

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Signal Processing

Citations in the corpus

Co-Separating Sounds of Visual Objects (2019, methodology): “While many methods assume audio captured by multiple microphones, some tackle the “blind” separation problem with singlechannel audio [22, 7, 48, 9], most recently with deep learning [20, 18, 44].”
A General Flexible Framework for the Handling of Prior Information in Audio Source Separation (2012, background): “Our proposition is complement ary to [9], [34] in that it accounts for temporal behaviour in the model structure itself in addition to possible priors on the model parameters.”
Nonnegative Matrix Factorization with the Itakura-Saito Divergence: With Application to Music Analysis (2009, methodology): “Pointing out the scale invariance of this measure, this work leads Virtanen (2007) to consider the IS divergence (but again without referring to it as such) for NMF in the context of singlechannel source separation, but the algorithm is applied to the magnitude spectra instead of the power spectra,…”
An Overview of Noise-Robust Automatic Speech Recognition (2014, methodology): “This is a source separation approach, and non-negative matrix factorization (NMF) [166] has been shown to be a very successful method [167], [168], and can directly bene ﬁ t noise-robust ASR [163]–[165], [169].”
Supervised and Unsupervised Speech Enhancement Using Nonnegative Matrix Factorization (2013, background): “, [24]) and that both of them are better than the results o f a state-of-the-art Wiener filter.”
Nonnegative signal factorization with learnt instrument models for sound source separation in close-microphone recordings (2013, background): “2 NMF background Factorization-based approaches have been intensively applied in signal processing applications, including singlechannel audio source separation [29,34-36].”
Speech Dereverberation Using Non-Negative Convolutive Transfer Function and Spectro-Temporal Modeling (2016, background): “Hence, since individual frequency bins are processed independently, these methods completely ignore the spectral structure of the speech signal.”
Direction of Arrival Based Spatial Covariance Model for Blind Sound Source Separation (2014, neutral): “More recently, methods based on finding spectrally redundant parts by non-negative matrix factorization (NMF) [12]– [14] have been proposed for separation of sound sources both with single [15], [16] and multichannel mixtures [17]–[20].”
Audio-Visual Scene Analysis with Self-Supervised Multisensory Features (2018, background): “Researchers have proposed many successful probabilistic approaches to this problem [30,31,32,33].”
Deep neural networks for single channel source separation (2013, methodology): “ex. Due to the linearity of the STFT, we have: Y(t;f) = S 1(t;f) + S 2(t;f); (1) where S 1(t;f) and S 2(t;f) are the unknown STFT of the ﬁrst and second sources in the mixed signal. In this framework [13, 24], the phase angles of the STFT were usually ignored. Hence, we can approximate the magnitude”
Determined Blind Source Separation Unifying Independent Vector Analysis and Nonnegative Matrix Factorization (2016, background): “On the other hand, for an underdetermined situation (number of microphones < number of sources) including monaural recording, nonnegative matrix factorization (NMF) [7], [8] has received much attention [9], [10].”
SpeakerBeam: Speaker Aware Neural Network for Target Speaker Extraction in Speech Mixtures (2019, methodology): “This problem has been studied in the past using methods, such as Computational auditory scene analysis [3], [4], Non-negative matrix factorization [5], [6]”
Singing Voice Separation with Deep U-Net Convolutional Networks (2017, positive): “Successful results have been achieved with non-negative matrix factorization [26, 30, 32], Bayesian methods [21], and the analysis of repeating structures [23].”
Single channel speech music separation using nonnegative matrix factorization and spectral masks (2011, methodology): “An unsupervised NMF with clustering was used in [4], to separate the mixed signal without any training data or any prior information about the underlying mixed signals.”
Multimodal Data Fusion: An Overview of Methods, Challenges, and Prospects (2015, background): “In such cases, the link is often defined as some correspondence between datasets that can be interpreted as similarity, smoothness or continuity [169].”
Self-supervised Audio-visual Co-segmentation (2019, methodology): “Semantic segmentation of images [2, 3] and sound source separation in audio [4, 5, 6, 7] are two important and popular tasks in the computer vision and computational audition communities.”
Exemplar-Based Sparse Representations for Noise Robust Automatic Speech Recognition (2011, methodology): “We use the generalized Kullback–Leibler (KL) divergence for (7) In source separation methods, the KL divergence has been found to produce better results than for example the Euclidean distance [12].”