Section: New Results

Emerging activities on compressive sensing, learning and inverse problems

Compressive sensing, acoustic wavefields, audio inpainting,

Nearfield acoustic holography (ECHANGE ANR project)

Participants : Rémi Gribonval, Nancy Bertin.

Main collaborations: Albert Cohen (Laboratoire Jacques-Louis Lions, Université Paris 6), Laurent Daudet, Gilles Chardon, François Ollivier, Antoine Peillot (Institut Jean Le Rond d'Alembert, Université Paris 6)

Compressed sensing is a rapidly emerging field which proposes a new approach to sample data far below the Nyquist rate when the sampled data admits a sparse approximation in some appropriate dictionary. The approach is supported by many theoretical results on the identification of sparse representations in overcomplete dictionaries, but many challenges remain open to determine its range of effective applicability. METISS has chosen to focus more specifically on the exploration of Compressed Sensing of Acoustic Wavefields, and we have set up the ANR collaborative project ECHANGE (ECHantillonnage Acoustique Nouvelle GEnération) which began in January 2009. Rémi Gribonval is the coordinator of the project.

In 2010, the activity on ECHANGE has concentrated on Nearfield acoustic holography (NAH), a technique aiming at reconstructing the operational deflection shapes of a vibrating structure, from the near sound field it generates. In this application scenario, the objective is either to improve the quality of the reconstruction (for a given number of sensors), or reduce the number of sensors, or both, by exploiting a sparsity hypothesis which helps regularizing the inverse problem involved.

Contributions of the team in this task spans: notations and model definitions, experimental setting design and implementation, choice of an adapted dictionary in which the sparsity hypothesis holds, improved acquisition strategies through pseudo-random sensor arrays and/or spatial multiplexing of the inputs, experimental study of robustness issues, and theoretical study of potential success guarantees based on the restricted isometry property (which revealed being not verified in our case, despite improved experimental performance).

A paper about robustness issues and spatial multiplexing (an alternative to building antennas with random sensor position) was published in GRETSI last year and as a journal paper this year [38] .

Sparse reconstruction for underwater acoustics (ECHANGE ANR project)

Participants : Rémi Gribonval, Nancy Bertin.

Main collaborations: Jacques Marchal, Pierre Cervenka (UPMC Univ Paris 06)

Underwater acoustic imaging is traditionally performed with beamforming: beams are formed at emission to insonify limited angular regions; beams are (synthetically) formed at reception to form the image. We proposed to exploit a natural sparsity prior to perform 3D underwater imaging using a newly built flexible-configuration sonar device. The computational challenges raised by the high-dimensionality of the problem were highlighted, and we described a strategy to overcome them. As a proof of concept, the proposed approach was used on real data acquired with the new sonar to obtain an image of an underwater target. We discussed the merits of the obtained image in comparison with standard beamforming, as well as the main challenges lying ahead, and the bottlenecks that will need to be solved before sparse methods can be fully exploited in the context of underwater compressed 3D sonar imaging. This work has been published in [61] and a journal paper is in preparation.

Audio inpainting (SMALL FET-Open project)

Participants : Rémi Gribonval, Nancy Bertin, Corentin Guichaoua.

Main collaborations: Amir Adler, Michael Elad (Computer Science Department, The Technion, Israel); Maria G. Jafari, Mark D. Plumbley (Centre for Digital Music, Department of Electronic Engineering, Queen Mary University of London, U.K.).

Inpainting is a particular kind of inverse problems that has been extensively addressed in the recent years in the field of image processing. It consists in reconstructing a set of missing pixels in an image based on the observation of the remaining pixels. Sparse representations have proved to be particularly appropriate to address this problem. However, inpainting audio data has never been defined as such so far.

METISS has initiated a series of works about audio inpainting, from its definition to methods to address it. This research has begun in the framework of the EU Framework 7 FET-Open project FP7-ICT-225913-SMALL (Sparse Models, Algorithms and Learning for Large-Scale data) which began in January 2009. Rémi Gribonval is the coordinator of the project. The research on audio inpainting has been conducted by Valentin Emiya in 2010 and 2011.

The contributions consist of:

  • defining audio inpainting as a general scheme where missing audio data must be estimated: it covers a number of existing audio processing tasks that have been addressed separately so far – click removal, declipping, packet loss concealment, unmasking in time-frequency;

  • proposing algorithms based on sparse representations for audio inpainting (based on Matching Pursuit and on l 1 minimization);

  • addressing the case of audio declipping (i.e. desaturation): thanks to the flexibility of our inpainting algorithms, they can be constrained so as to include the structure of signals due to clipping in the objective to optimize. The resulting performance are significantly improved. This work will appear as a journal paper [33] .

Current and future works deal with developping advanced sparse decomposition for audio inpainting, including several forms of structured sparsity (e.g. temporal and multichannel joint-sparsity), dictionary learning for inpainting, and several applicative scenarios (declipping, time-frequency inpainting).

Blind Calibration of Compressive Sensing systems

Participants : Rémi Gribonval, Cagdas Bilen.

Main collaborations: Gilles Chardon, Laurent Daudet (Institut Langevin), Gilles Puy (EPFL)

We consider the problem of calibrating a compressed sensing measurement system under the assumption that the decalibration consists in unknown gains on each measure. We focus on blind calibration, using measures performed on a few unknown (but sparse) signals. A naive formulation of this blind calibration problem, using l1 minimization, is reminiscent of blind source separation and dictionary learning, which are known to be highly non-convex and riddled with local minima. In the considered context, we show that in fact this formulation can be exactly expressed as a convex optimization problem, and can be solved using off-the-shelf algorithms. Numerical simulations demonstrate the effectiveness of the approach even for highly uncalibrated measures, when a sufficient number of (unknown, but sparse) calibrating signals is provided. We observe that the success/failure of the approach seems to obey sharp phase transitions. This work has been published at ICASSP 2012 [54] , and an extension dealing with the problem of phase-only decalibration, using techniques revolving around low-rank matrix recovery, has been submitted to ICASSP 2013. A journal version is in preparation.

Compressive Gaussian Mixture estimation

Participants : Rémi Gribonval, Anthony Bourrier.

Main collaborations: Gilles Blanchard (University of Potsdam), Patrick Perez (Technicolor R&D, FR)

When fitting a probability model to voluminous data, memory and computational time can become prohibitive. In this paper, we pro- pose a framework aimed at fitting a mixture of isotropic Gaussians to data vectors by computing a low-dimensional sketch of the data. The sketch represents empirical moments of the underlying probability distribution. Deriving a reconstruction algorithm by analogy with compressive sensing, we experimentally show that it is possible to precisely estimate the mixture parameters provided that the sketch is large enough. Our algorithm provides good reconstruction and scales to higher dimensions than previous probability mixture estimation algorithms, while consuming less memory in the case of numerous data. It also provides a privacy-preserving data analysis tool, since the sketch does not disclose information about individual datum it is based on. This work has been submitted for publication at ICASSP 2013.

Nearest neighbor search for arbitrary kernels with explicit embeddings

Participants : Rémi Gribonval, Anthony Bourrier.

Main collaborations: Hervé Jégou (TEX-MEX team), Patrick Perez (Technicolor R&D, FR)

Many algorithms have been proposed to handle efficient search in large databases for simple metrics such as the Euclidean distance. However, few approaches apply to more sophisticated Positive Semi-Definite (PSD) kernels. In this document, we propose for such kernels to use the concept of explicit embedding and to cast the search problem into a Euclidean space. We first describe an exact nearest neighbor search technique which relies on bounds on the approximation of the kernel. We show that, in the case of SIFT descriptors, one can retrieve the nearest neighbor with probability 1 by computing only a fraction of the costly kernels between the query and the database vectors. We then propose to combine explicit embedding with a recent Euclidean approximate nearest neighbor search method and show that it leads to significant improvements with respect to the state-of-the-art methods which rely on an implicit embedding. The database vectors being indexed by short codes, the approach is shown to scale to a dataset comprising 200 million vectors on a commodity server. This work has been submitted for journal publication [74]