logo cnrs  logo oca  logo_uns  

Cédric Févotte

CNRS research scientist

Laboratoire Lagrange
(CNRS, Observatoire de la Côte d'Azur & Université de Nice Sophia Antipolis)

Parc Valrose
06108 Nice cedex 2

Tel: +33 4 92 07 63 48

Email:  cfevotte -at- unice -dot- fr

link to publications 
(with sound files and code, if any)


selected recent talks

  • Maximum marginal likelihood estimation for nonnegative dictionary learning, Xerox Research Center Europe, Grenoble, Jan. 2014. slides
  • Non-negative dynamical system with application to speech and audio, ICASSP, Vancouver CA, June 2013. slides
  • Itakura-Saito nonnegative matrix factorization and friends for music signal decompositionModern Methods of Time-Frequency Analysis Semester, Erwin Schroedinger Institute, Vienna, Dec. 2012. slides


Born 1977 in Northeastern France, grew up in Tunisia, Senegal and Madagascar.

Current position

Previous positions

Education & academic qualifications

Visiting positions
Select professional activities

Misc : my photoblog.

research interests

statistical signal processing and machine learning, for denoising, source separation, inverse problems, in particular with audio applications, and including statistical estimation, Bayesian inference, optimization, stochastic simulation, variational approximations, latent variable models, independent component analysis, sparse component analysis, nonnegative matrix factorization, time-frequency representations.

my most cited papers, as of Google Scholar.


original audio    restored audio

(use stereo headphones for optimal rendering)

  • one of our chord recognition systems developed with Laurent Oudre and Yves Grenier is used in the music search engine and mashup application MuMa, developed by Exalead

software and other resources

MATLAB code associated with many of my papers can be found on my publications page.

I am with Emmanuel Vincent and Rémi Gribonval one of the creators of the widely used blind source separation evaluation Matlab toolbox BSS_EVAL.

I run the Latent Variable Analysis and Signal Separation LVA/ICA mailing list.

research projects

Active projects
  • CNRS MASTODONS project DISPLAY (Distributed processing for very large arrays in radioastronomy)
Past projects I was involved in
  • Lead scientist of ANR young researcher project TANGERINE (Theory and Application of Nonnegative Matrix Factorization)
  • ANR project SARAH ("Standardisation du Remastering Audio Haute-définition").
  • European Comission Network of Excellence K-Space ("Knowledge Space of Semantic Inference for Automatic Annotation and Retrieval of Multimedia Content").
  • European Commission Specific Targeted Research Project MEMORIES ("Design for an audio semantic indexation system allowing information retrieval for the access to archive content").
  • EPSRC Research Network ICArn on Blind Source Separation and Independent Component Analysis.
  • European Comission Research Training Network HASSIP ("Harmonic Analysis and Statistics for SIgnal Processing").
  • European Commission Research Network MOUMIR ("MOdels for Unified Multimedia Information Retrieval").
  • Action Jeunes Chercheurs CNRS-GdR ISIS "Ressources pour la séparation de sources audio".
  • European Shared Cost Project NEW ("Neuromuscular assessment in the Elderly Worker").


Recent collaborators :
Past and present PhD students :
  • Nicolas Seichepine (since 2012; co-supervised with Slim Essid and Olivier Cappé)
Past and present visitors :
  • Dennis Sun (Stanford University), summer 2013.
  • Brian King (University of Washington), spring 2012.

Some events I (co-)organised :

Un rapport (in French) :

P. Aimé and C. Févotte, La simplification administrative de la gestion des unités de recherche, Rapport de l'Inspection Générale de l'Éducation Nationale et de la Recherche (IGAENR), no 2008-089, 2008.

Last Update: Feb. 15th, 2015