Convolutional neural networks (CNN) are often computationally demanding for mobile devices.Offloading some computation lowers this burden: initial convolutional layers are processed on a smartphone, the resulting high dimensional features are transmitted, and latter layers are processed in the cloud/edge/another device.To improve this process, we propose Dynamic Switch, a convolutional subnetwork enabling anywhere splittable CNNs with multirate feature compression using a single set of network parameters.We achieve 90% feature compression with at most 3% accuracy loss for MobileNet and MSDNet on ImageNet dataset and at most 4.58% on CIFAR100 dataset with MSDNet, ResNet-18, Mo-bileNet/MobileNetv2 and ShuffleNet/ShuffleNetv2.

Neural style transfer, allowing to apply the artistic style of one image to another, has become one of the most widely showcased computer vision applications shortly after its introduction. In contrast, related tasks in the music audio domain remained, until recently, largely untackled. While several style conversion methods tailored to musical signals have been proposed, most lack the ‘one-shot’ capability of classical image style transfer algorithms. On the other hand, the results of existing one-shot audio style transfer methods on musical inputs are not as compelling. In this work, we are specifically interested in the problem of one-shot timbre transfer. We present a novel method for this task, based on an extension of the vector-quantized variational autoencoder (VQ-VAE), along with a simple self-supervised learning strategy designed to obtain disentangled representations of timbre and pitch. We evaluate the method using a set of objective metrics and show that it is able to outperform selected baselines.

Audio-visual (AV) representation learning is an important task from the perspective of designing machines with the ability to understand complex events. To this end, we propose a novel multimodal framework that instantiates multiple instance learning. Specifically, we develop methods that identify events and localize corresponding AV cues in unconstrained videos. Importantly, this is done using weak labels where only video-level event labels are known without any information about their location in time. We show that the learnt representations are useful for performing several tasks such as event/object classification, audio event detection, audio source separation and visual object localization. An important feature of our method is its capacity to learn from unsynchronized audio-visual events. We also demonstrate our framework's ability to separate out the audio source of interest through a novel use of nonnegative matrix factorization. State-of-the-art classification results, with a F1-score of 65.0, are achieved on DCASE 2017 smart cars challenge data with promising generalization to diverse object types such as musical instruments. Visualizations of localized visual regions and audio segments substantiate our system's efficacy, especially when dealing with noisy situations where modality-specific cues appear asynchronously.

“Style transfer” among images has recently emerged as a very active research topic, fuelled by the power of convolution neural networks (CNNs), and has become fast a very popular technology in social media. This paper investigates the analogous problem in the audio domain: How to transfer the style of a reference audio signal to a target audio content? We propose a flexible framework for the task, which uses a sound texture model to extract statistics characterizing the reference audio style, followed by an optimization-based audio texture synthesis to modify the target content. In contrast to mainstream optimization-based visual transfer method, the proposed process is initialized by the target content instead of random noise and the optimized loss is only about texture, not structure. These differences proved key for audio style transfer in our experiments. In order to extract features of interest, we investigate different architectures, whether pre-trained on other tasks, as done in image style transfer, or engineered based on the human auditory system. Experimental results on different types of audio signal confirm the potential of the proposed approach.

In this paper we tackle the problem of single channel audio source separation driven by descriptors of the sounding object's motion. As opposed to previous approaches, motion is included as a soft-coupling constraint within the nonnegative matrix factorization framework. The proposed method is applied to a multimodal dataset of instruments in string quartet performance recordings where bow motion information is used for separation of string instruments. We show that the approach offers better source separation result than an audio-based baseline and the state-of-the-art multimodal-based approaches on these very challenging music mixtures.

We propose a novel informed source separation method for audio object coding based on a recent sampling theory for smooth signals on graphs. Assuming that only one source is active at each time-frequency point, we compute an ideal map indicating which source is active at each time-frequency point at the encoder. This map is then sampled with a compressive graph signal sampling strategy that guarantees accurate and stable recovery at the decoder. The graph is built using feature vectors, computed using non-negative matrix factorization, that allows us to connect similar source activations in the time-frequency plane. We show that the proposed approach performs better than state-of-the-art methods at low bitrate.

This paper addresses the challenging problem of single-channel audio source separation. We introduce a novel user-guided framework where source models that govern the separation process are learned on-the-fly from audio examples retrieved online. The user only provides the search keywords that describe the sources in the mixture. In this framework, the generic spectral characteristics of each source are modeled by a universal sound class model learned from the retrieved examples via nonnegative matrix factorization. We propose several group sparsity-inducing constraints in order to efficiently exploit a relevant subset of the universal model adapted to the mixture to be separated. We then derive the corresponding multiplicative update rules for parameter estimation. Separation results obtained from automated and user tests on mixtures containing various types of sounds confirm the effectiveness of the proposed framework.

Speech enhancement and separation are core problems in audio signal processing, with commercial applications in devices as diverse as mobile phones, conference call systems, hands-free systems, or hearing aids. In addition, they are crucial preprocessing steps for noise-robust automatic speech and speaker recognition. Many devices now have two to eight microphones. The enhancement and separation capabilities offered by these multichannel interfaces are usually greater than those of single-channel interfaces. Research in speech enhancement and separation has followed two convergent paths, starting with microphone array processing and blind source separation, respectively. These communities are now strongly interrelated and routinely borrow ideas from each other. Yet, a comprehensive overview of the common foundations and the differences between these approaches is lacking at present. In this paper, we propose to fill this gap by analyzing a large number of established and recent techniques according to four transverse axes: 1) the acoustic impulse response model, 2) the spatial filter design criterion, 3) the parameter estimation algorithm, and 4) optional postfiltering. We conclude this overview paper by providing a list of software and data resources and by discussing perspectives and future trends in the field.

Nonnegative matrix or tensor factorization is a very popular approach for audio source separation. One important problem in nonnegative tensor factorization (NTF) in the context of user-guided audio source separation is the necessity to manually assign the NTF components to audio sources in order to be able to enforce prior information on the sources during the estimation process. In this paper, two new approaches to NTF based source separation are proposed, which do not require any manual component assignment to the sources, but estimate the underlying assignment automatically. Both algorithms use the prior information on the source samples in the estimation process along with either a limit on the minimum number of components each source uses or with a restriction that each component is used by sparse number of sources. The proposed methods are shown to outperform the classic approach with a manual distribution of the components equally among the sources.

Audio declipping consists in recovering so-called clipped audio samples that are set to a maximum / minimum threshold. Many different approaches were proposed to solve this problem in case of singlechannel (mono) recordings. However, while most of audio recordings are multichannel nowadays, there is no method designed specifically for multichannel audio declipping, where the inter-channel correlations may be efficiently exploited for a better declipping result. In this work we propose for the first time such a multichannel audio declipping method. Our method is based on representing a multichannel audio recording as a convolutive mixture of several audio sources, and on modeling the source power spectrograms and mixing filters by nonnegative tensor factorization model and full-rank covariance matrices, respectively. A generalized expectation-maximization algorithm is proposed to estimate model parameters. It is shown experimentally that the proposed multichannel audio de-clipping algorithm outperforms in average and in most cases a state-of-the-art single-channel declipping algorithm applied to each channel independently.

We consider dictionary-based signal decompositions with group sparsity, a variant of structured sparsity. We point out that the group sparsity-inducing constraint alone may not be sufficient in some cases when we know that some bigger groups or so-called supergroups cannot vanish completely. To deal with this problem we introduce the notion of relative group sparsity preventing the supergroups from vanishing. In this paper we formulate practical criteria and algorithms for relative group sparsity as applied to non-negative matrix factorization and investigate its potential benefit within the on-the-fly audio source separation framework we recently introduced. Experimental evaluation shows that the proposed relative group sparsity leads to performance improvement over group sparsity in both supervised and semi-supervised on-the-fly audio source separation settings.

Though audio source separation offers a wide range of applications in audio enhancement and post-production, its performance has yet to reach the satisfactory especially for single-channel mixtures with limited training data. In this paper we present a novel interactive source separation framework that allows end-users to provide feedback at each separation step so as to gradually improve the result. For this purpose, a prototype graphical user interface (GUI) is developed to help users annotating time-frequency regions where a source can be labeled as either active, inactive, or well-separated within the displayed spectrogram. This user feedback information, which is partially new with respect to the state-of-the-art annotations, is then taken into account in a proposed uncertainty-based learning algorithm to constraint the source estimates in next separation step. The considered framework is based on non-negative matrix factorization and is shown to be effective even without using any isolated training data.

Automatic extraction of face tracks is a key component of systems that analyse people in audio-visual content such as TV programs and movies. Due to the lack of properly annotated content of this type, popular algorithms for extracting face tracks have not been fully assessed in the literature. We introduce and make publicly available a new dataset, based on the full annotation of a feature movie, to help fill this gap. We show in particular that, thanks to this dataset, state-of-art tracking metrics can now be exploited to evaluate face tracks used by, e.g., automatic character naming systems. We conduct such an evaluation on different variants of a novel system that we introduce as a generalization of existing ones.

Informed source separation (ISS) aims at reliably recovering sources from a mixture. To this purpose, it relies on the assumption that the original sources are available during an encoding stage. Given both sources and mixture, a side-information may be computed and transmitted along with the mixture, whereas the original sources are not available any longer. During a decoding stage, both mixture and side-information are processed to recover the sources. ISS is motivated by a number of specific applications including active listening and remixing of music, karaoke, audio gaming, etc. Most ISS techniques proposed so far rely on a source separation strategy and cannot achieve better results than oracle estimators. In this study, we introduce Coding-based ISS (CISS) and draw the connection between ISS and source coding. CISS amounts to encode the sources using not only a model as in source coding but also the observation of the mixture. This strategy has several advantages over conventional ISS methods. First, it can reach any quality, provided sufficient bandwidth is available as in source coding. Second, it makes use of the mixture in order to reduce the bitrate required to transmit the sources, as in classical ISS. Furthermore, we introduce Nonnegative Tensor Factorization as a very efficient model for CISS and report rate-distortion results that strongly outperform the state of the art.

The coding of audio-visual signals is generally based on different paradigms for high and low rates. At high rates the signal is approximated directly and at low rates only signal features are transmitted. The recently introduced distribution preserving quantization (DPQ) paradigm provides a seamless transition between these two regimes. In this paper we present a simplified scheme that preserves the power spectral density (PSD) rather than the probability distribution. In a practical system the PSD must be estimated. We show that both forward adaptive and backward adaptive PSD estimation are possible. Our experimental results confirm that preservation of PSD at finite precision leads to a unified coding paradigm that provides effective coding at both high and low rates. An audio coding application shows the perceptual benefits of PSD preserving quantization.

Most audio source separation methods are developed for a particular scenario characterized by the number of sources and channels and the characteristics of the sources and the mixing process. In this paper, we introduce a general audio source separation framework based on a library of structured source models that enable the incorporation of prior knowledge about each source via user-specifiable constraints. While this framework generalizes several existing audio source separation methods, it also allows to imagine and implement new efficient methods that were not yet reported in the literature. We first introduce the framework by describing the model structure and constraints, explaining its generality, and summarizing its algorithmic implementation using a generalized expectation-maximization algorithm. Finally, we illustrate the above-mentioned capabilities of the framework by applying it in several new and existing configurations to different source separation problems. We have released a software tool named Flexible Audio Source Separation Toolbox (FASST) implementing a baseline version of the framework in Matlab.

This paper deals with the localization of multiple sources from two-channel mixtures recorded in a reverberant environment. We introduce new angular spectrum-based methods relying on the signal-to-noise ratio (SNR) to estimate the time difference of arrival (TDOA) of each source. We propose and compare five ways of estimating the SNR in each time-frequency point and in each direction, using beamforming techniques and statistical models. Large-scale evaluation considering a high number of situations shows the effectiveness of the proposed approach compared to state-of-the-art angular spectrum-based techniques.

Separating multiple tracks from professionally produced music recordings (PPMRs) is still a challenging problem. We address this task with a user-guided approach in which the separation system is provided segmental information indicating the time activations of the particular instruments to separate. This information may typically be retrieved from manual annotation. We use a so-called multichannel nonnegative tensor factorization (NTF) model, in which the original sources are observed through a multichannel convolutive mixture and in which the source power spectrograms are jointly modeled by a 3-valence (time/frequency/source) tensor. Our user-guided separation method produced competitive results at the 2010 Signal Separation Evaluation Campaign, with sufficient quality for real-world music editing applications.

Using high-rate theory approximations we introduce flexible practical quantizers based on possibly non-Gaussian models in both the constrained resolution (CR) and the constrained entropy cases. We derive model estimation criteria optimizing asymptotic (with increasing rate) quantizer performance. We show that in the CR case the optimal criterion is different from the maximum likelihood criterion commonly used for that purpose and introduce a new criterion that we call constrained resolution minimum description length (CR-MDL). We apply these principles to the generalized Gaussian scaled mixture model, which is accurate for many real-world signals. We provide an explanation of the reason why the CR-MDL improves quantization performance in the CR case and show that CR-MDL can compensate for a possible mismatch between model and data distribution. Thus, this criterion is of a great interest for practical applications. Our experiments apply the new quantization method to controllable artificial data and to the commonly used modulated lapped transform representation of audio signals. We show that both the CR-MDL criterion and a non-Gaussian modeling have significant advantages.

We consider inference in a general data-driven object-based model of multichannel audio data, assumed generated as a possibly underdetermined convolutive mixture of source signals. We work in the short-time Fourier transform (STFT) domain, where convolution is routinely approximated as linear instantaneous mixing in each frequency band. Each source STFT is given a model inspired from nonnegative matrix factorization (NMF) with the Itakura-Saito divergence, which underlies a statistical model of superimposed Gaussian components. We address estimation of the mixing and source parameters using two methods. The first one consists of maximizing the exact joint likelihood of the multichannel data using an expectation-maximization (EM) algorithm. The second method consists of maximizing the sum of individual likelihoods of all channels using a multiplicative update algorithm inspired from NMF methodology. Our decomposition algorithms are applied to stereo audio source separation in various settings, covering blind and supervised separation, music and speech sources, synthetic instantaneous and convolutive mixtures, as well as professionally produced music recordings. Our EM method produces competitive results with respect to state-of-the-art as illustrated on two tasks from the international Signal Separation Evaluation Campaign (SiSEC 2008).

We consider inference in a general data-driven object-based model of multichannel audio data, assumed generated as a possibly under-determined convolutive mixture of source signals. Each source is given a model inspired from nonnegative matrix factorization (NMF) with the Itakura-Saito divergence, which underlies a statistical model of superimposed Gaussian components. We address estimation of the mixing and source parameters using two methods. The first one consists of maximizing the exact joint likelihood of the multichannel data using an expectation-maximization algorithm. The second method consists of maximizing the sum of individual likelihoods of all channels using a multiplicative update algorithm inspired from NMF methodology. Our decomposition algorithms were applied to stereo music and assessed in terms of blind source separation performance.

We address optimal model estimation for model-based vector quantization for both the constrained resolution (CR) and constrained entropy (CE) cases. To this purpose we derive under high-rate (HR) theory assumptions the rate-distortion (RD) relations for these two quantization scenarios assuming a Gaussian model. Based on the RD relations we show that the maximum likelihood (ML) criterion leads to optimal performance for CE quantization, but not for CR quantization. We introduce a new model estimation criterion for CR quantization that is optimal (under HR theory assumptions) in terms of the RD relation. Our experiments confirm that the proposed criterion for model identification outperforms the ML criterion for a range of conditions.

Probabilistic approaches can offer satisfactory solutions to source separation with a single channel, provided that the models of the sources match accurately the statistical properties of the mixed signals. However, it is not always possible to train such models. To overcome this problem, we propose to resort to an adaptation scheme for adjusting the source models with respect to the actual properties of the signals observed in the mix. In this paper, we introduce a general formalism for source model adaptation which is expressed in the framework of Bayesian models. Particular cases of the proposed approach are then investigated experimentally on the problem of separating voice from music in popular songs. The obtained results show that an adaptation scheme can improve consistently and significantly the separation performance in comparison with nonadapted models.