Using graphs to model irregular information domains is an effective approach to deal with some of the intricacies of contemporary (network) data. A key aspect is how the data, represented as graph signals, depend on the topology of the graph. Widely-used approaches assume that the observed signals can be viewed as outputs of graph filters (i.e., polynomials of a matrix representation of the graph) whose inputs have a particular structure. Diffused graph signals, which correspond to an originally sparse (node-localized) signal percolated through the graph via filtering, fall into this class. In that context, this paper deals with the problem of jointly identifying graph filters and separating their (sparse) input signals from a mixture of diffused graph signals, thus generalizing to the graph signal processing framework the classical blind demixing (blind source separation) of temporal and spatial signals. We first consider the scenario where the supporting graphs are different across the signals, providing a theorem for demixing feasibility along with probabilistic bounds on successful recovery. Additionally, an analysis of the degenerate problem of demixing with a single graph is also presented. Numerical experiments with synthetic and real-world graphs empirically illustrating the main theoretical findings close the paper.
Tackling unsupervised source separation jointly with an additional inverse problem such as deconvolution is central for the analysis of multi-wavelength data. This becomes highly challenging when applied to large data sampled on the sphere such as those provided by wide-field observations in astrophysics, whose analysis requires the design of dedicated robust and yet effective algorithms. We therefore investigate a new joint deconvolution/sparse blind source separation method dedicated for data sampled on the sphere, coined SDecGMCA. It is based on a projected alternate least-squares minimization scheme, whose accuracy is proved to strongly rely on some regularization scheme in the present joint deconvolution/blind source separation setting. To this end, a regularization strategy is introduced that allows designing a new robust and effective algorithm, which is key to analyze large spherical data. Numerical experiments are carried out on toy examples and realistic astronomical data.
The mismatch between the numerical and actual nonlinear models is a challenge to nonlinear acoustic echo cancellation (NAEC) when the nonlinear adaptive filter is utilized. To alleviate this problem, we combine a basis-generic expansion of the memoryless nonlinearity into semi-blind source separation (SBSS). By regarding all the basis functions of the far-end input signal as the known equivalent reference signals, an SBSS updating algorithm is derived following the constrained scaled natural gradient strategy. Unlike the commonly utilized adaptive algorithm, the proposed SBSS is based on the independence between the near-end signal and the reference signals, and is less sensitive to the mismatch of nonlinearity between the numerical and actual models. Experimental results show that the proposed method outperforms conventional methods in terms of echo return loss enhancement (ERLE) and near-end speech quality evaluated by perceptual evaluation of speech quality (PESQ) and short-time objective intelligibility (STOI).
We propose a new algorithm for blind source separation (BSS) using independent vector analysis (IVA). This is an improvement over the popular auxiliary function based IVA (AuxIVA) with iterative projection (IP) or iterative source steering (ISS). We introduce iterative projection with adjustment (IPA), where we update one demixing filter and jointly adjust all the other sources along its current direction. Each update involves solving a non-convex minimization problem that we term log-quadratically penalized quadratic minimization (LQPQM), that we think is of interest beyond this work. In the general case, we show that its global minimum corresponds to the largest root of a univariate function, reminiscent of modified eigenvalue problems. We propose a simple procedure based on Newton-Raphson to efficiently compute it. Numerical experiments demonstrate the effectiveness of the proposed method. First, we show that it efficiently decreases the value of the surrogate function. In further experiments on synthetic mixtures, we study the probability of finding the true demixing matrix and convergence speed. We show that the proposed method combines high success rate and fast convergence. Finally, we validate the performance on a reverberant blind speech separation task. We find that all the AuxIVA-based methods perform similarly in terms of acoustic BSS metrics. However, AuxIVA-IPA converges faster. We measure up to 8.5 times speed-up in terms of runtime compared to the next best AuxIVA-based method, depending on the number of channels and the signal-to-noise ratio (SNR).
Due to age-related hearing loss, elderly people often struggle with following the language on TV. Because they form an increasing part of the audience, this problem will become even more important in the future and needs to be addressed by research and development. Object-based audio is a promising approach to solve this issue as it offers the possibility of customizable dialogue enhancement (DE). For this thesis an Adjustment / Satisfaction Test (A/ST) was conducted to evaluate the preferred loudness difference (LD) between speech and background in people above 65. Two different types of DE were tested: DE with separately available audio components (speech and background) and DE with components created by blind source separation (BSS). The preferred LDs compared to the original, differences of the preferred LDs between the two DE methods and the listener satisfaction were tested. It was observed that the preferred LDs were larger than the original LDs, that customizable DE increases listener satisfaction and that the two DE methods performed comparably well in terms of preferred LD and listener satisfaction. Based on the results, it can be assumed that elderly viewers above 65 will benefit equally from user-adjustable DE by available components and by dialogue separation.
Multichannel audio blind source separation (BSS) in the determined situation (the number of microphones is equal to that of the sources), or determined BSS, is performed by multichannel linear filtering in the time-frequency domain to handle the convolutive mixing process. Ordinarily, the filter treats each frequency independently, which causes the well-known permutation problem, i.e., the problem of how to align the frequency-wise filters so that each separated component is correctly assigned to the corresponding sources. In this paper, it is shown that the general property of the time-frequency-domain representation called spectrogram consistency can be an assistant for solving the permutation problem.
This paper proposes harmonic vector analysis (HVA) based on a general algorithmic framework of audio blind source separation (BSS) that is also presented in this paper. BSS for a convolutive audio mixture is usually performed by multichannel linear filtering when the numbers of microphones and sources are equal (determined situation). This paper addresses such determined BSS based on batch processing. To estimate the demixing filters, effective modeling of the source signals is important. One successful example is independent vector analysis (IVA) that models the signals via co-occurrence among the frequency components in each source. To give more freedom to the source modeling, a general framework of determined BSS is presented in this paper. It is based on the plug-and-play scheme using a primal-dual splitting algorithm and enables us to model the source signals implicitly through a time-frequency mask. By using the proposed framework, determined BSS algorithms can be developed by designing masks that enhance the source signals. As an example of its application, we propose HVA by defining a time-frequency mask that enhances the harmonic structure of audio signals via sparsity of cepstrum. The experiments showed that HVA outperforms IVA and independent low-rank matrix analysis (ILRMA) for both speech and music signals. A MATLAB code is provided along with the paper for a reference ( https://doi.org/10.24433/CO.9507820.v1 ).
Algorithms for Blind Source Separation (BSS) of acoustic signals require efficient and fast converging optimization strategies to adapt to nonstationary signal statistics and time-varying acoustic scenarios. In this paper, we derive fast converging update rules from a negentropy perspective, which are based on the Majorize-Minimize (MM) principle and eigenvalue decomposition. The presented update rules are shown to outperform competing state-of-the-art methods in terms of convergence speed at a comparable runtime due to the restriction to unitary demixing matrices. This is demonstrated by experiments with recorded real-world data.
Signal separation and extraction are important tasks for devices recording audio signals in real environments which, aside from the desired sources, often contain several interfering sources such as background noise or concurrent speakers. Blind Source Separation (BSS) provides a powerful approach to address such problems. However, BSS algorithms typically treat all sources equally and do not resolve uncertainty regarding the ordering of the separated signals at the output of the algorithm, i.e., the outer permutation problem. This paper addresses this problem by incorporating prior knowledge into the adaptation of the demixing filters, e.g., the position of the sources, in a Bayesian framework. We focus here on methods based on Independent Vector Analysis (IVA) as it elegantly and successfully deals with the internal permutation problem. By including a background model, i.e., a model for sources we are not interested to separate, we enable the algorithm to extract the sources of interest in overdetermined and underdetermined scenarios at a low computational complexity. The proposed framework allows to incorporate prior knowledge about the demixing filters in a generic way and unifies several known and newly proposed algorithms using a Bayesian view. For all algorithmic variants, we provide efficient update rules based on the iterative projection principle. The performance of a large variety of representative algorithmic variants, including very recent algorithms, is compared using measured room impulse responses.
We present a Maximum A Posteriori (MAP) derivation of the Independent Vector Analysis (IVA) algorithm, a blind source separation algorithm, by incorporating a prior over the demixing matrices, relying on a free-field model. In this way, the outer permutation ambiguity of IVA is avoided. The resulting MAP optimization problem is solved by deriving majorize-minimize update rules to achieve convergence speed comparable to the well-known auxiliary function IVA algorithm. The performance of the proposed algorithm is investigated and compared to a benchmark algorithm using real measurements.
Blind source separation (BSS) techniques aims at joint estimation of source signals and a mixing matrix from observations of mixtures. This paper addresses a doubly nonstationary BSS problem, where the mixing matrix is time dependent and sources are nonstationary, more precisely deformed stationary signals, following the model of [1]. An algorithm for joint BSS and estimation of stationarity-breaking deformations and spectra is introduced, that exploits suitable approximations for the behavior of the wavelet transform of such nonstationary signals. The performance of the approach is evaluated on numerical simulations, and compared with other nonstationary BSS algorithms.
This article introduces quaternion non-negative matrix factorization (QNMF), which generalizes the usual non-negative matrix factorization (NMF) to the case of polarized signals. Polarization information is represented by Stokes parameters, a set of 4 energetic parameters widely used in polarimetric imaging. QNMF relies on two key ingredients: (i) the algebraic representation of Stokes parameters thanks to quaternions and (ii) the exploitation of physical constraints on Stokes parameters. These constraints generalize non-negativity to the case of polarized signals, encoding positive semi-definiteness of the covariance matrix associated which each source. Uniqueness conditions for QNMF are presented. Remarkably, they encompass known sufficient uniqueness conditions from NMF. Meanwhile, QNMF further relaxes NMF uniqueness conditions requiring sources to exhibit a certain zero-pattern, by leveraging the complete polarization information. We introduce a simple yet efficient algorithm called quaternion alternating least squares (QALS) to solve the QNMF problem in practice. Closed-form quaternion updates are obtained using the recently introduced generalized HR calculus. Numerical experiments on synthetic data demonstrate the relevance of the approach. QNMF defines a promising generic low-rank approximation tool to handle polarization, notably for blind source separation problems arising in imaging applications.
Images Stacks as Parametric Surfaces (ISPS) is a powerful model that was originally proposed for image registration. Being closely related to mutual information (MI) - the most classic similarity measure for image registration, ISPS works well across different categories of registration problems. The Signals as Parametric Curves (SPC) model is derived from ISPS extended to 1-dimensional signals. Blind Source Separation (BSS) is a classic problem in signal processing, where Independent Component Analysis (ICA) based approaches are popular and effective. Since MI plays an important role in ICA, based on the close relationship with MI, we apply SPC model to BSS in this paper, and propose a group of geometrical objective functions that are simple yet powerful, and serve as replacements of original MI-based objective functions. Motivated by the geometrical objective functions, we also propose a second-order-statistics approach, FT-PCA. Both geometrical objective functions and FT-PCA consider signals as functions instead of stochastic processes, make use of derivative information of signals, and do not rely on the independence assumption. In this paper, we discuss the reasonability of the assumptions of geometrical objective functions and FT-PCA, and show their effectiveness by synthetic experiments, comparing with other previous classic approaches.
Blind source separation (BSS) is addressed, using a novel data-driven approach, based on a well-established probabilistic model. The proposed method is specifically designed for separation of multichannel audio mixtures. The algorithm relies on spectral decomposition of the correlation matrix between different time frames. The probabilistic model implies that the column space of the correlation matrix is spanned by the probabilities of the various speakers across time. The number of speakers is recovered by the eigenvalue decay, and the eigenvectors form a simplex of the speakers' probabilities. Time frames dominated by each of the speakers are identified exploiting convex geometry tools on the recovered simplex. The mixing acoustic channels are estimated utilizing the identified sets of frames, and a linear umixing is performed to extract the individual speakers. The derived simplexes are visually demonstrated for mixtures of 2, 3 and 4 speakers. We also conduct a comprehensive experimental study, showing high separation capabilities in various reverberation conditions.
The TRINICON ('Triple-N ICA for convolutive mixtures') framework is an effective blind signal separation (BSS) method for separating sound sources from convolutive mixtures. It makes full use of the non-whiteness, non-stationarity and non-Gaussianity properties of the source signals and can be implemented either in time domain or in frequency domain, avoiding the notorious internal permutation problem. It usually has best performance when the sources are continuously mixed. In this paper, the offline dual-channel frequency domain TRINICON implementation for sparsely mixed signals is investigated, and a multi-source activity detection is proposed to locate the active period of each source, based on which the filter updating strategy is regularized to improve the separation performance. The objective metric provided by the BSSEVAL toolkit is utilized to evaluate the performance of the proposed scheme.
Previous theoretical studies on the interaction of excitatory and inhibitory neurons proposed to model this cortical microcircuit motif as a so-called Winner-Take-All (WTA) circuit. A recent modeling study however found that the WTA model is not adequate for data-based softer forms of divisive inhibition as found in a microcircuit motif in cortical layer 2/3. We investigate here through theoretical analysis the role of such softer divisive inhibition for the emergence of computational operations and neural codes under spike-timing dependent plasticity (STDP). We show that in contrast to WTA models - where the network activity has been interpreted as probabilistic inference in a generative mixture distribution - this network dynamics approximates inference in a noisy-OR-like generative model that explains the network input based on multiple hidden causes. Furthermore, we show that STDP optimizes the parameters of this model by approximating online the expectation maximization (EM) algorithm. This theoretical analysis corroborates a preceding modelling study which suggested that the learning dynamics of this layer 2/3 microcircuit motif extracts a specific modular representation of the input and thus performs blind source separation on the input statistics.
Non-invasive fetal electrocardiogram (FECG) is used to monitor the electrical pulse of the fetal heart. Decomposing the FECG signal from maternal ECG (MECG) is a blind source separation problem, which is hard due to the low amplitude of FECG, the overlap of R waves, and the potential exposure to noise from different sources. Traditional decomposition techniques, such as adaptive filters, require tuning, alignment, or pre-configuration, such as modeling the noise or desired signal. to map MECG to FECG efficiently. The high correlation between maternal and fetal ECG parts decreases the performance of convolution layers. Therefore, the masking region of interest using the attention mechanism is performed for improving signal generators' precision. The sine activation function is also used since it could retain more details when converting two signal domains. Three available datasets from the Physionet, including A&D FECG, NI-FECG, and NI-FECG challenge, and one synthetic dataset using FECGSYN toolbox, are used to evaluate the performance. The proposed method could map abdominal MECG to scalp FECG with an average 98% R-Square [CI 95%: 97%, 99%] as the goodness of fit on A&D FECG dataset. Moreover, it achieved 99.7 % F1-score [CI 95%: 97.8-99.9], 99.6% F1-score [CI 95%: 98.2%, 99.9%] and 99.3% F1-score [CI 95%: 95.3%, 99.9%] for fetal QRS detection on, A&D FECG, NI-FECG and NI-FECG challenge datasets, respectively. These results are comparable to the state-of-the-art; thus, the proposed algorithm has the potential of being used for high-performance signal-to-signal conversion.
In this work, we consider the problem of blind source separation (BSS) by departing from the usual linear model and focusing on the linear-quadratic (LQ) model. We propose two provably robust and computationally tractable algorithms to tackle this problem under separability assumptions which require the sources to appear as samples in the data set. The first algorithm generalizes the successive nonnegative projection algorithm (SNPA), designed for linear BSS, and is referred to as SNPALQ. By explicitly modeling the product terms inherent to the LQ model along the iterations of the SNPA scheme, the nonlinear contributions of the mixing are mitigated, thus improving the separation quality. SNPALQ is shown to be able to recover the ground truth factors that generated the data, even in the presence of noise. The second algorithm is a brute-force (BF) algorithm, which is used as a post-processing step for SNPALQ. It enables to discard the spurious (mixed) samples extracted by SNPALQ, thus broadening its applicability. The BF is in turn shown to be robust to noise under easier-to-check and milder conditions than SNPALQ. We show that SNPALQ with and without the BF postprocessing is relevant in realistic numerical experiments.
In this work, we consider the application of model-based deep learning in nonlinear principal component analysis (PCA). Inspired by the deep unfolding methodology, we propose a task-based deep learning approach, referred to as Deep-RLS, that unfolds the iterations of the well-known recursive least squares (RLS) algorithm into the layers of a deep neural network in order to perform nonlinear PCA. In particular, we formulate the nonlinear PCA for the blind source separation (BSS) problem and show through numerical analysis that Deep-RLS results in a significant improvement in the accuracy of recovering the source signals in BSS when compared to the traditional RLS algorithm.
A Sturm-Liouville problem ($\lambda wy=(ry')'+qy$) is singular if its domain is unbounded or if $r$ or $w$ vanish at the boundary. Then it is difficult to tell whether profound results from regular Sturm-Liouville theory apply. Existing criteria are often difficult to apply, e.g. because they are formulated in terms of the solution function. We study the special case that the potential $q$ is zero under Neumann boundary conditions and give simple and explicit criteria, solely in terms of the coefficient functions, to assess whether various properties of the regular case apply. Specifically, these properties are discreteness of the spectrum (BD), self-adjointness, oscillation ($i$th solution has $i$ zeros) and that the $i$th eigenvalue equals the SFA delta value (the total energy) of the $i$th solution. We further prove that stationary points of each solution strictly interlace with its zeros (in singular or regular case, regardless of the boundary condition, for zero potential or if $q < \lambda w$ everywhere). If $\frac{r}{w}$ is bounded and of bounded variation, the criterion simplifies to requiring $\frac{|w'|}{w} \to \infty$ at singular boundary points. This research is motivated by Slow Feature Analysis (SFA), a data processing algorithm that extracts the slowest uncorrelated signals from a high-dimensional input signal and has notable success in computer vision, computational neuroscience and blind source separation. From [Sprekeler et al., 2014] it is known that for an important class of scenarios (statistically independent input), an analytic formulation of SFA reduces to a Sturm-Liouville problem with zero potential and Neumann boundary conditions. So far, the mathematical SFA theory has only considered the regular case, except for a special case that is solved by Hermite Polynomials. This work generalizes SFA theory to the singular case, i.e. open-space scenarios.