Opendata, web and dolomites

FACTORY SIGNED

New paradigms for latent factor estimation

Total Cost €

0

EC-Contrib. €

0

Partnership

0

Views

0

 FACTORY project word cloud

Explore the words cloud of the FACTORY project. It provides you a very rough idea of what is the project "FACTORY" about.

columns    methodology    multimodal    made    pays    latent    object    describes    first    samples    world    explore    reconsidered    semantic    optimal    off    completion    speech    approximation    appears    settings    enhancement    discovery    yields    nature    rank    retrieval    mining    transform    paradigm    recommendation    dominant    enhanced    loosely    raw    break    assumption    collection    transforms    inducing    clustering    geometry    dimensionality    too    unmixing    remastering    push    recurring    matrix    dictionary    deterministic    cosmic    lfe    music    frontiers    computed    thirdly    shown    approximate    patterns    marginal    names    constraints    estimation    soft    societal    gaussian    performance    usually    proportions    rarely    single    earth    remote    domain    shelf    factorisation    complies    audio    diverse    ubiquitous    signal    form    learnt    limit    entails    data    observation    practical    coding    difficult    models    preprocess    learning    statistical    paradigms    setting    sensing   

Project "FACTORY" data sheet

The following table provides information about the project.

Coordinator
CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE CNRS 

Organization address
address: RUE MICHEL ANGE 3
city: PARIS
postcode: 75794
website: www.cnrs.fr

contact info
title: n.a.
name: n.a.
surname: n.a.
function: n.a.
email: n.a.
telephone: n.a.
fax: n.a.

 Coordinator Country France [FR]
 Project website http://projectfactory.irit.fr/
 Total cost 1˙931˙776 €
 EC max contribution 1˙931˙776 € (100%)
 Programme 1. H2020-EU.1.1. (EXCELLENT SCIENCE - European Research Council (ERC))
 Code Call ERC-2015-CoG
 Funding Scheme ERC-COG
 Starting year 2016
 Duration (year-month-day) from 2016-09-01   to  2021-08-31

 Partnership

Take a look of project's partnership.

# participants  country  role  EC contrib. [€] 
1    CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE CNRS FR (PARIS) coordinator 1˙931˙776.00

Map

 Project objective

Data is often available in matrix form, in which columns are samples, and processing of such data often entails finding an approximate factorisation of the matrix in two factors. The first factor yields recurring patterns characteristic of the data. The second factor describes in which proportions each data sample is made of these patterns. Latent factor estimation (LFE) is the problem of finding such a factorisation, usually under given constraints. LFE appears under other domain-specific names such as dictionary learning, low-rank approximation, factor analysis or latent semantic analysis. It is used for tasks such as dimensionality reduction, unmixing, soft clustering, coding or matrix completion in very diverse fields.

In this project, I propose to explore three new paradigms that push the frontiers of traditional LFE. First, I want to break beyond the ubiquitous Gaussian assumption, a practical choice that too rarely complies with the nature and geometry of the data. Estimation in non-Gaussian models is more difficult, but recent work in audio and text processing has shown that it pays off in practice. Second, in traditional settings the data matrix is often a collection of features computed from raw data. These features are computed with generic off-the-shelf transforms that loosely preprocess the data, setting a limit to performance. I propose a new paradigm in which an optimal low-rank inducing transform is learnt together with the factors in a single step. Thirdly, I show that the dominant deterministic approach to LFE should be reconsidered and I propose a novel statistical estimation paradigm, based on the marginal likelihood, with enhanced capabilities. The new methodology is applied to real-world problems with societal impact in audio signal processing (speech enhancement, music remastering), remote sensing (Earth observation, cosmic object discovery) and data mining (multimodal information retrieval, user recommendation).

 Publications

year authors and title journal last update
List of publications.
2016 R. Flamary, C. Févotte, N. Courty, and V. Emiya
Optimal spectral transportation with application to music transcription
published pages: , ISSN: , DOI:
Advances in Neural Information Processing Systems (NIPS) 2019-06-18
2018 D. Fagot, H. Wendt, and C. Févotte
Nonnegative matrix factorization with transform learning
published pages: , ISSN: , DOI:
Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2019-06-18
2018 A. Ozerov, C. Févotte, and E. Vincent
An introduction to multichannel NMF for audio source separation
published pages: , ISSN: , DOI:
Audio Source Separation 2019-06-18
2018 C. Févotte, E. Vincent, and A. Ozerov
Single-channel audio source separation with NMF: divergences, constraints and algorithms
published pages: , ISSN: , DOI:
Audio Source Separation 2019-06-18
2018 C. Févotte, P. Smaragdis, N. Mohammadiha, and G. Mysore
Temporal extensions of nonnegative matrix factorization
published pages: , ISSN: , DOI:
Audio Source Separation and Speech Enhancement 2019-06-18
2018 L. Filstroff, A. Lumbreras, and C. Févotte
Closed-form marginal likelihood in Gamma-Poisson matrix factorization
published pages: , ISSN: , DOI:
Proc. International Conference on Machine Learning (ICML) 2019-05-27
2018 O. Gouvert, T. Oberlin, and C. Févotte
Matrix co-factorization for cold-start recommendation
published pages: , ISSN: , DOI:
Proc. International Society for Music Information Retrieval Conference (ISMIR) 2019-05-27
2019 Y. C. Cavalcanti, T. Oberlin, N. Dobigeon, C. Févotte, S. Stute, and C. Tauber
Factor analysis of dynamic PET images: beyond Gaussian noise
published pages: , ISSN: 0278-0062, DOI:
IEEE Transactions on Medical Imaging 2019-05-15
2019 R. Xia, V. Y. F. Tan, L. Filstroff and C. Févotte
A ranking model motivated by nonnegative matrix factorization with applications to tennis tournaments
published pages: , ISSN: , DOI:
arXiv 2019-05-15
2018 H. Wendt, D. Fagot, and C. Févotte
Jacobi algorithm for nonnegative matrix factorization with transform learning
published pages: , ISSN: , DOI:
Proc. European Signal Processing Conference (EUSIPCO) 2019-04-18
2018 A. Lumbreras, L. Filstroff, and C. Févotte
Bayesian mean-parameterized nonnegative binary matrix factorization
published pages: , ISSN: , DOI:
arXiv 2019-04-18
2019 P. Ablin, D. Fagot, H. Wendt, A. Gramfort, and C. Févotte
A quasi-Newton algorithm on the orthogonal manifold for NMF with transform learning
published pages: , ISSN: , DOI:
Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2019-04-18
2019 Y. C. Cavalcanti, T. Oberlin, N. Dobigeon, C. Févotte, S. Stute, and C. Tauber
Unmixing dynamic PET images: Combining spatial heterogeneity and non-gaussian noise
published pages: , ISSN: , DOI:
Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2019-04-18
2018 O. Gouvert, T. Oberlin, and C. Févotte
Negative binomial matrix factorization for recommender systems
published pages: , ISSN: , DOI:
arXiv 2019-04-18
2019 D. Fagot, H. Wendt, C. Févotte, and P. Smaragdis
Majorization-minimization algorithms for convolutive NMF with the beta-divergence
published pages: , ISSN: , DOI:
Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2019-04-18
2018 C. Févotte and M. Kowalski
Estimation with low-rank time-frequency synthesis models
published pages: , ISSN: 1053-587X, DOI:
IEEE Transactions on Signal Processing 2019-04-18

Are you the coordinator (or a participant) of this project? Plaese send me more information about the "FACTORY" project.

For instance: the website url (it has not provided by EU-opendata yet), the logo, a more detailed description of the project (in plain text as a rtf file or a word file), some pictures (as picture files, not embedded into any word file), twitter account, linkedin page, etc.

Send me an  email (fabio@fabiodisconzi.com) and I put them in your project's page as son as possible.

Thanks. And then put a link of this page into your project's website.

The information about "FACTORY" are provided by the European Opendata Portal: CORDIS opendata.

More projects from the same programme (H2020-EU.1.1.)

MOCHA (2019)

Understanding and leveraging ‘moments of change’ for pro-environmental behaviour shifts

Read More  

UTOPEST (2019)

Unified Theory of Efficient Optimization and Estimation

Read More  

CohoSing (2019)

Cohomology and Singularities

Read More