Recent advances in deep learning have significantly improved multichannel speech enhancement algorithms, yet conventional training loss functions such as the scale-invariant signal-to-distortion ratio (SDR) may fail to preserve fine-grained spectral cues essential for phoneme intelligibility. In this work, we propose perceptually-informed variants of the SDR loss, formulated in the time-frequency domain and modulated by frequency-dependent weighting schemes. These weights are designed to emphasize time-frequency regions where speech is prominent or where the interfering noise is particularly strong. We investigate both fixed and adaptive strategies, including ANSI band-importance weights, spectral magnitude-based weighting, and dynamic weighting based on the relative amount of speech and noise. We train the FaSNet multichannel speech enhancement model using these various losses. Experimental results show that while standard metrics such as the SDR are only marginally improved, their perceptual frequency-weighted counterparts exhibit a more substantial improvement. Besides, spectral and phoneme-level analysis indicates better consonant reconstruction, which points to a better preservation of certain acoustic cues.

Audio inpainting, i.e., the task of restoring missing or occluded audio signal samples, usually relies on sparse representations or autoregressive modeling. In this paper, we propose to structure the spectrogram with nonnegative matrix factorization (NMF) in a probabilistic framework. First, we treat the missing samples as latent variables, and derive two expectation–maximization algorithms for estimating the parameters of the model, depending on whether we formulate the problem in the time- or time-frequency domain. Then, we treat the missing samples as parameters, and we address this novel problem by deriving an alternating minimization scheme. We assess the potential of these algorithms for the task of restoring short- to middle-length gaps in music signals. Experiments reveal great convergence properties of the proposed methods, as well as competitive performance when compared to state-of-the-art audio inpainting techniques.

Structuring the latent space in probabilistic deep generative models, e.g., variational autoencoders (VAEs), is important to yield more expressive models and interpretable representations, and to avoid overfitting.One way to achieve this objective is to impose a sparsity constraint on the latent variables, e.g., via a Laplace prior.However, such approaches usually complicate the training phase, and they sacrifice the reconstruction quality to promote sparsity.In this paper, we propose a simple yet effective methodology to structure the latent space via a sparsity-promoting dictionary model, which assumes that each latent code can be written as a sparse linear combination of a dictionary's columns.In particular, we leverage a computationally efficient and tuning-free method, which relies on a zeromean Gaussian latent prior with learnable variances.We derive a variational inference scheme to train the model.Experiments on speech generative modeling demonstrate the advantage of the proposed approach over competing techniques, since it promotes sparsity while not deteriorating the output speech quality.

This paper tackles the problem of decomposing binary data using matrix factorization. We consider the family of mean-parametrized Bernoulli models, a class of generative models that are well suited for modeling binary data and enables interpretability of the factors. We factorize the Bernoulli parameter and consider an additional Beta prior on one of the factors to further improve the model’s expressive power. While similar models have been proposed in the literature, they only exploit the Beta prior as a proxy to ensure a valid Bernoulli parameter in a Bayesian setting; in practice it reduces to a uniform or uninformative prior. Besides, estimation in these models has focused on costly Bayesian inference. In this paper, we propose a simple yet very efficient majorization-minimization algorithm for maximum a posteriori estimation. Our approach leverages the Beta prior whose parameters can be tuned to improve performance in matrix completion tasks. Experiments conducted on three public binary datasets show that our approach offers an excellent trade-off between prediction performance, computational complexity, and interpretability.

This paper considers the phase retrieval (PR) problem, which aims to reconstruct a signal from phaseless measurements such as magnitude or power spectrograms. PR is generally handled as a minimization problem involving a quadratic loss. Recent works have considered alternative discrepancy measures, such as the Bregman divergences, but it is still challenging to tailor the optimal loss for a given setting. In this paper we propose a novel strategy to automatically learn the optimal metric for PR. We unfold a recently introduced ADMM algorithm into a neural network, and we emphasize that the information about the loss used to formulate the PR problem is conveyed by the proximity operator involved in the ADMM updates. Therefore, we replace this proximity operator with trainable activation functions: learning these in a supervised setting is then equivalent to learning an optimal metric for PR. Experiments conducted with speech signals show that our approach outperforms the baseline ADMM, using a light and interpretable neural architecture.

State-of-the-art music recommendation systems are based on collaborative filtering, which predicts a user’s interest from his listening habits and similarities with other users’ profiles. These approaches are agnostic to the song content, and therefore face the cold-start problem: they cannot recommend novel songs without listening history. To tackle this issue, content-aware recommendation incorporates information about the songs that can be used for recommending new items. Most methods falling in this category exploit either user-annotated tags, acoustic features or deeply-learned features. Consequently, these content features do not have a clear musical meaning, thus they are not necessarily relevant from a musical preference perspective. In this work, we propose instead to leverage a model of musical preference which originates from the field of music psychology. From low-level acoustic features we extract three factors (arousal, valence and depth), which have been shown appropriate for describing musical taste. Then we integrate those into a collaborative filtering framework for content-aware music recommendation. Experiments conducted on large-scale data show that this approach is able to address the cold-start problem, while using a compact and meaningful set of musical features.

Time-frequency audio source separation is usually achieved by estimating the short-time Fourier transform (STFT) magnitude of each source, and then applying a phase recovery algorithm to retrieve time-domain signals. In particular, the multiple input spectrogram inversion (MISI) algorithm has shown good performance in several recent works. This algorithm minimizes a quadratic reconstruction error between magnitude spectrograms. However, this loss does not properly account for some perceptual properties of audio, and alternative discrepancy measures such as beta-divergences have been preferred in many settings. In this paper, we propose to reformulate phase recovery in audio source separation as a minimization problem involving Bregman divergences. To optimize the resulting objective, we derive a projected gradient descent algorithm. Experiments conducted on a speech enhancement task show that this approach out-performs MISI for several alternative losses, which highlights their relevance for audio source separation applications.

Audio source separation is usually achieved by estimating the short-time Fourier transform (STFT) magnitude of each source, and then applying a spectrogram inversion algorithm to retrieve time-domain signals. In particular, the multiple input spectrogram inversion (MISI) algorithm has been exploited successfully in several recent works. However, this algorithm suffers from two drawbacks, which we address in this letter. First, it has originally been introduced in a heuristic fashion: we propose here a rigorous optimization framework in which MISI is derived, thus proving the convergence of this algorithm. Besides, while MISI operates offline, we propose here an online version of MISI called oMISI, which is suitable for low-latency source separation, an important requirement for e.g., hearing aids applications. oMISI also allows one to use alternative phase initialization schemes exploiting the temporal structure of audio signals. Experiments conducted on a speech separation task show that oMISI performs as well as its offline counterpart, thus demonstrating its potential for real-time source separation.

This paper introduces a phase-aware probabilistic model for audio source separation. Classical source models in the short-time Fourier transform domain use circularly-symmetric Gaussian or Poisson random variables. This is equivalent to assuming that the phase of each source is uniformly distributed, which is not suitable for exploiting the underlying structure of the phase. Drawing on preliminary works, we introduce here a Bayesian anisotropic Gaussian source model in which the phase is no longer uniform. Such a model permits us to favor a phase value that originates from a signal model through a Markov chain prior structure. The variance of the latent variables are structured with nonnegative matrix factorization (NMF). The resulting model is called complex Itakura-Saito NMF (ISNMF) since it generalizes the ISNMF model to the case of nonisotropic variables. It combines the advantages of ISNMF, which uses a distortion measure adapted to audio and yields a set of estimates which preserve the overall energy of the mixture, and of complex NMF, which enables one to account for some phase constraints. We derive a generalized expectation-maximization algorithm to estimate the model parameters. Experiments conducted on a musical source separation task in a semiinformed setting show that the proposed approach outperforms state-of-the-art phase-aware separation techniques.

In audio source separation applications, it is common to model the sources as circular-symmetric Gaussian random variables, which is equivalent to assuming that the phase of each source is uniformly distributed. In this paper, we introduce an anisotropic Gaussian source model in which both the magnitude and phase parameters are modeled as random variables. In such a model, it becomes possible to promote a phase value that originates from a signal model and to adjust the relative importance of this underlying model-based phase constraint. We conduct Bayesian inference of the model through the derivation of an expectation-maximization algorithm for estimating the parameters. Experiments conducted on realistic music songs for a monaural source separation task, in an scenario where the variance parameters are assumed known, show that the proposed approach outperforms state-of-the-art techniques.

S.332-336

International audience

For audio source separation applications, it is common to estimate the magnitude of the short-time Fourier transform (STFT) of each source. In order to further synthesize time-domain signals, it is necessary to recover the phase of the corresponding complex-valued STFT. Most authors in this field choose a Wiener-like filtering approach, which boils down to use the phase of the original mixture. In this paper, a different standpoint is adopted. Many music events are partially composed of slowly varying sinusoids and the STFT phase increment over time of those frequency components takes a specific form. This allows phase recovery by an unwrapping technique once a short-term frequency estimate has been obtained. Herein, a novel iterative source separation procedure is proposed that builds upon these results. It consists in minimizing the mixing error by means of the auxiliary function method. This procedure is initialized by exploiting the unwrapping technique in order to generate estimates that benefit from a temporal continuity property. Experiments conducted on realistic music pieces show that, given accurate magnitude estimates, this procedure outperforms the state-of-the-art consistent Wiener filter.

Phase reconstruction of complex components in the time-frequency domain is a challenging but necessary task for audio source separation. While traditional approaches do not exploit phase constraints that originate from signal modeling, some prior information about the phase can be obtained from sinusoidal modeling. In this paper, we introduce a probabilistic mixture model which allows us to incorporate such phase priors within a source separation framework. While the magnitudes are estimated beforehand, the phases are modeled by Von Mises random variables whose location parameters are the phase priors. We then approximate this non-tractable model by an anisotropic Gaussian model, in which the phase dependencies are preserved. This enables us to derive an MMSE estimator of the sources which optimally combines Wiener filtering and prior phase estimates. Experimental results highlight the potential of incorporating phase priors into mixture models for separating overlapping components in complex audio mixtures.

Nonnegative Matrix Factorization (NMF) is a powerful tool for decomposing mixtures of audio signals in the Time-Frequency (TF) domain. In the source separation framework, the phase recovery for each extracted component is necessary for synthesizing time-domain signals. The Complex NMF (CNMF) model aims to jointly estimate the spectrogram and the phase of the sources, but requires to constrain the phase in order to produce satisfactory sounding results. We propose to incorporate phase constraints based on signal models within the CNMF framework: a phase unwrapping constraint that enforces a form of temporal coherence, and a constraint based on the repetition of audio events, which models the phases of the sources within onset frames. We also provide an algorithm for estimating the model parameters. The experimental results highlight the interest of including such constraints in the CNMF framework for separating overlapping components in complex audio mixtures.

In this paper we present a novel source separation method aiming to overcome the difficulty of modelling non-stationary signals. The method can be applied to mixtures of musical instruments with frequency and/or amplitude modulation, e.g. typically caused by vibrato. It is based on a signal representation that divides the complex spectrogram into a grid of patches of arbitrary size. These complex patches are then processed by a two-dimensional discrete Fourier transform, forming a tensor representation which reveals spectral and temporal modulation textures. Our representation can be seen as an alternative to modulation transforms computed on magnitude spectrograms. An adapted factorization model allows to decompose different time-varying harmonic sources based on their particular common modulation profile: hence the name Common Fate Model. The method is evaluated on musical instrument mixtures playing the same fundamental frequency (unison), showing improvement over other state-of-the-art methods.

Nonnegative Matrix Factorization (NMF) is a powerful tool for decomposing mixtures of audio signals in the Time-Frequency (TF) domain. In applications such as source separation, the phase recovery for each extracted component is a major issue since it often leads to audible artifacts. In this paper, we present a methodology for evaluating various NMF-based source separation techniques involving phase reconstruction. For each model considered, a comparison between two approaches (blind separation without prior information and oracle separation with supervised model learning) is performed, in order to inquire about the room for improvement for the estimation methods. Experimental results show that the High Resolution NMF (HRNMF) model is particularly promising, because it is able to take phases and correlations over time into account with a great expressive power.