QuickSearch:   Number of matching entries: 0.

Search Settings

    AuthorTitleYearJournal/Proceedings/InstitutionBibTeX typeURL
    R. Flamary, C. Févotte, N. Courty & V. Emiya Optimal spectral transportation with application to music transcription 2016 Advances in Neural Information Processing Systems (NIPS)   inproceedings [PDF] [CODE]  
    BibTeX:
    
    @inproceedings{Flamary2016,
      author = {R.~Flamary and C.~Févotte and N.~Courty and V.~Emiya} ,
      title = {Optimal spectral transportation with application to music transcription},
      booktitle = {Advances in Neural Information Processing Systems (NIPS)},
      year = {2016},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/nips16.pdf}
    }
    
    R. Hanon, V. Emiya & C. Févotte Convex nonnegative matrix factorization with missing data 2016 Proc.~IEEE International Workshop on Machine Learning for Signal Processing (MLSP)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{Hanon2016,
      author = {R.~Hanon and V.~Emiya and C.~Févotte} ,
      title = {Convex nonnegative matrix factorization with missing data},
      booktitle = {Proc.~IEEE International Workshop on Machine Learning for Signal Processing (MLSP)},
      year = {2016},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/mlsp16.pdf}
    }
    
    R. Hanon, V. Emiya & C. Févotte Factorisation archétypale en matrices non-négatives avec données manquantes 2016 Proc.~Conférence francophone sur l'Apprentissage Automatique (CAP)   inproceedings  
    BibTeX:
    
    @inproceedings{Hanon2016a,
      author = {R.~Hanon and V.~Emiya and C.~Févotte} ,
      title = {Factorisation archétypale en matrices non-négatives avec données manquantes},
      booktitle = {Proc.~Conférence francophone sur l'Apprentissage Automatique (CAP)},
      year = {2016}
    }
    
    C. Févotte & J. Grivel Apprentissage pour la synthèse visuelle guidée par un flux audio et application dans le champ de l'art contemporain 2015 Proc.~Colloque GRETSI sur le Traitement du Signal et des Images   inproceedings [PDF] [DEMO]  
    BibTeX:
    
    @inproceedings{Fevotte2015,
      author = {C.~Févotte and J.~Grivel} ,
      title = {Apprentissage pour la synthèse visuelle guidée par un flux audio et application dans le champ de l'art contemporain},
      booktitle = {Proc.~Colloque GRETSI sur le Traitement du Signal et des Images},
      address = {Lyon, France},
      month = {Sep.},
      year = {2015},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/gretsi15a.pdf}
    }
    
    C. Févotte & M. Kowalski Hybrid sparse and low-rank time-frequency signal decomposition 2015 Proc.~European Signal Processing Conference (EUSIPCO)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{Fevotte2015a,
      author = {C.~Févotte and M.~Kowalski} ,
      title = {Hybrid sparse and low-rank time-frequency signal decomposition},
      booktitle = {Proc.~European Signal Processing Conference (EUSIPCO)},
      address = {Nice, France},
      month = {Sep.},
      year = {2015},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/eusipco15.pdf}
    }
    
    R. Hanon, P. Borgnat, C. Févotte, P. Flandrin & C. Robardet Factorisation de réseaux temporels : étude des rythmes hebdomadaires du système Vélo'v 2015 Proc.~Colloque GRETSI sur le Traitement du Signal et des Images   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{Hanon2015,
      author = {R.~Hanon and P.~Borgnat and C.~Févotte and P.~Flandrin and C.~Robardet} ,
      title = {Factorisation de réseaux temporels : étude des rythmes hebdomadaires du système Vélo'v},
      booktitle = {Proc.~Colloque GRETSI sur le Traitement du Signal et des Images},
      address = {Lyon, France},
      month = {Sep.},
      year = {2015},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/gretsi15b.pdf}
    }
    
    C. Févotte & N. Dobigeon Nonlinear hyperspectral unmixing with robust nonnegative matrix factorization 2015 IEEE Transactions on Image Processing   article [PDF] [DOI] [CODE]  
    BibTeX:
    
    @article{tip15,
      author = {Févotte, C. and Dobigeon, N.} ,
      title = {Nonlinear hyperspectral unmixing with robust nonnegative matrix factorization},
      journal = {IEEE Transactions on Image Processing},
      month = {Dec.},
      year = {2015},
      volume = {24},
      number = {12},
      pages = {4810-4819},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/tip2015.pdf},
      doi = {http://dx.doi.org/10.1109/TIP.2015.2468177}
    }
    
    C. Févotte Contributions à la factorisation en matrices non-négatives 2014 Mémoire d'Habilitation à Diriger des Recherches. Université Nice Sophia Antipolis   misc [PDF] [SLIDES]  
    BibTeX:
    
    @misc{Fevotte2014a,
      author = {C.~Févotte} ,
      title = {Contributions à la factorisation en matrices non-négatives},
      month = {Oct.},
      year = {2014},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/theses/hdr_fevotte.pdf},
      slides = {https://www.irit.fr/~Cedric.Fevotte/publications/theses/hdr_fevotte.pdf}
    }
    
    D.L. Sun & C. Févotte Alternating direction method of multipliers for non-negative matrix factorization with the beta-divergence 2014 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF] [CODE]  
    BibTeX:
    
    @inproceedings{icassp14a,
      author = {D.~L.~Sun and C.~Févotte} ,
      title = {Alternating direction method of multipliers for non-negative matrix factorization with the beta-divergence},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Florence, Italy},
      month = {May},
      year = {2014},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp14a.pdf}
    }
    
    N. Seichepine, S. Essid, C. Févotte & O. Cappé Piecewise constant nonnegative matrix factorization 2014 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{icassp14b,
      author = {N.~Seichepine and S.~Essid and C.~Févotte and O.~Cappé} ,
      title = {Piecewise constant nonnegative matrix factorization},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Florence, Italy},
      month = {May},
      year = {2014},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp14b.pdf}
    }
    
    C. Févotte & M. Kowalski Low-rank time-frequency synthesis 2014 Advances in Neural Information Processing Systems (NIPS)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{nips14,
      author = {C.~Févotte and M.~Kowalski} ,
      title = {Low-rank time-frequency synthesis},
      booktitle = {Advances in Neural Information Processing Systems (NIPS)},
      month = {Dec.},
      year = {2014},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/nips14.pdf}
    }
    
    N. Seichepine, S. Essid, C. Févotte & O. Cappé Soft nonnegative matrix co-factorization 2014 IEEE Transactions on Signal Processing   article [PDF]  
    BibTeX:
    
    @article{Seichepine2014,
      author = {N.~Seichepine and S.~Essid and C.~Févotte and O.~Cappé} ,
      title = {Soft nonnegative matrix co-factorization},
      journal = {IEEE Transactions on Signal Processing},
      month = {Nov.},
      year = {2014},
      volume = {62},
      number = {22},
      pages = {5940--5949},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/tsp2014.pdf}
    }
    
    P. Smaragdis, C. Févotte, G. Mysore, N. Mohammadiha & M. Hoffman Static and dynamic source separation using nonnegative factorizations: A unified view 2014 IEEE Signal Processing Magazine   article [PDF] [DOI]  
    BibTeX:
    
    @article{Smaragdis2014,
      author = {P.~Smaragdis and C.~Févotte and G.~Mysore and N.~Mohammadiha and M.~Hoffman} ,
      title = {Static and dynamic source separation using nonnegative factorizations: A unified view},
      journal = {IEEE Signal Processing Magazine},
      month = {May},
      year = {2014},
      volume = {31},
      number = {3},
      pages = {66-75},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/spm2014.pdf},
      doi = {http://dx.doi.org/10.1109/MSP.2013.2297715}
    }
    
    V.Y.F. Tan & C. Févotte Automatic relevance determination in nonnegative matrix factorization with the beta-divergence 2013 IEEE Transactions on Pattern Analysis and Machine Intelligence   article [PDF] [CODE]  
    Abstract: This paper addresses the problem of estimating the latent dimensionality in nonnegative matrix factorization (NMF). The estimation is done via automatic relevance determination (ARD). Uncovering the model order is important as it is necessary to strike the right balance between data fidelity and overfitting. We propose a Bayesian model for NMF and two families of algorithms known as l1-ARD and l2-ARD, each assuming different priors on the basis elements and the activation coefficients. The proposed algorithms leverage on the recent algorithmic advances in NMF with the beta-divergence using majorization-minimization (MM) methods. They are parametrized by the shape parameter beta which governs the assumption on the noise statistics. We show by deriving various auxiliary functions that the cost functions of the algorithms decrease monotonically to a local minimum. We also describe a heuristic to select the hyperparameters based on the method of moments. We demonstrate the efficacy and robustness of our algorithms by performing extensive experiments on synthetic data, the swimmer dataset, a music decomposition example and a stock price prediction task.
    BibTeX:
    
    @article{ardnmfj,
      author = {V.~Y.~F.~Tan and C.~Févotte} ,
      title = {Automatic relevance determination in nonnegative matrix factorization with the beta-divergence},
      journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
      month = {July},
      year = {2013},
      volume = {35},
      number = {7},
      pages = {1592 -- 1605},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/pami13_ardnmf.pdf}
    }
    
    N. Seichepine, S. Essid, C. Févotte & O. Cappé Co-factorisation douce en matrices non-négatives. Application au regroupement multimodal de locuteurs 2013 Proc.~Colloque GRETSI sur le Traitement du Signal et des Images   inproceedings  
    BibTeX:
    
    @inproceedings{gretsi13,
      author = {N.~Seichepine and S.~Essid and C.~Févotte and O.~Cappé} ,
      title = {Co-factorisation douce en matrices non-négatives. Application au regroupement multimodal de locuteurs},
      booktitle = {Proc.~Colloque GRETSI sur le Traitement du Signal et des Images},
      address = {Brest, France},
      month = {Sep.},
      year = {2013}
    }
    
    C. Févotte, J. Le~Roux & J.R. Hershey Non-negative dynamical system with application to speech and audio 2013 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF] [CODE] [DEMO]  
    BibTeX:
    
    @inproceedings{icassp13a,
      author = {C.~Févotte and J.~Le~Roux and J.~R.~Hershey} ,
      title = {Non-negative dynamical system with application to speech and audio},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Vancouver, Canada},
      month = {May},
      year = {2013},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp13a.pdf}
    }
    
    N. Seichepine, S. Essid, C. Févotte & O. Cappé Soft nonnegative matrix co-factorization with application to multimodal speaker diarization 2013 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{icassp13b,
      author = {N.~Seichepine and S.~Essid and C.~Févotte and O.~Cappé} ,
      title = {Soft nonnegative matrix co-factorization with application to multimodal speaker diarization},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Vancouver, Canada},
      month = {May},
      year = {2013},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp13b.pdf}
    }
    
    J. Le Roux, C. Févotte & J.R. Hershey A new non-negative dynamical system for speech and audio modeling 2013 Proc.~Acoustical Society of Japan Spring Meeting   inproceedings  
    BibTeX:
    
    @inproceedings{LeRoux2013,
      author = {J.~Le Roux and C.~Févotte and J.~R.~Hershey} ,
      title = {A new non-negative dynamical system for speech and audio modeling},
      booktitle = {Proc.~Acoustical Society of Japan Spring Meeting},
      month = {Mar.},
      year = {2013}
    }
    
    S. Essid & C. Févotte Smooth nonnegative matrix factorization for unsupervised audiovisual document structuring 2013 IEEE Transactions on Multimedia   article [PDF] [DOI]  
    Abstract: This paper introduces a new paradigm for unsupervised audiovisual document structuring. In this paradigm, a novel Nonnegative Matrix Factorization (NMF) algorithm is applied on histograms of counts (relating to a bag of features representation of the content) to jointly discover latent structuring patterns and their activations in time. Our NMF variant employs the Kullback-Leibler divergence as a cost function and imposes a temporal smoothness constraint to the activations. It is solved for using a majorization-minimization technique. The approach proposed is meant to be generic and is particularly well suited to applications where the structuring patterns may overlap in time. As such, it is evaluated on a person-oriented video structuring task, using a challenging database of political debate videos. Our results outperform reference results obtained by a method using Hidden Markov Models. Further, we show the potential that our general approach has for audio speaker diarization.
    BibTeX:
    
    @article{nmfvideostruct,
      author = {S.~Essid and C.~Févotte} ,
      title = {Smooth nonnegative matrix factorization for unsupervised audiovisual document structuring},
      journal = {IEEE Transactions on Multimedia},
      month = {Feb.},
      year = {2013},
      volume = {15},
      number = {2},
      pages = {415--425},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_multimedia_smoothnmf.pdf},
      doi = {http://dx.doi.org/10.1109/TMM.2012.2228474}
    }
    
    N. Dobigeon & C. Févotte Robust nonnegative matrix factorization for nonlinear unmixing of hyperspectral images 2013 Proc.~IEEE Workshop Hyperspectral image and signal processing: Evolution in remote sensing (WHISPERS)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{whispers13,
      author = {Dobigeon, N. and Févotte, C.} ,
      title = {Robust nonnegative matrix factorization for nonlinear unmixing of hyperspectral images},
      booktitle = {Proc.~IEEE Workshop Hyperspectral image and signal processing: Evolution in remote sensing (WHISPERS)},
      address = {Gainesville, Florida},
      month = {June},
      year = {2013},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/whispers13.pdf}
    }
    
    S. Essid & C. Févotte Decomposing the video editing structure of a talk-show using nonnegative matrix factorization 2012 Proc. IEEE International Conference on Image Processing (ICIP)   inproceedings  
    BibTeX:
    
    @inproceedings{icip12,
      author = {S.~Essid and C.~Févotte} ,
      title = {Decomposing the video editing structure of a talk-show using nonnegative matrix factorization},
      booktitle = {Proc. IEEE International Conference on Image Processing (ICIP)},
      address = {Orlando, Florida},
      month = {Sep.},
      year = {2012}
    }
    
    A. Lefèvre, F. Bach & C. Févotte Semi-supervised NMF with time-frequency annotations for single-channel source separation 2012 Proc.~International Society for Music Information Retrieval Conference (ISMIR)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{ismir12a,
      author = {A.~Lefèvre and F.~Bach and C.~Févotte} ,
      title = {Semi-supervised NMF with time-frequency annotations for single-channel source separation},
      booktitle = {Proc.~International Society for Music Information Retrieval Conference (ISMIR)},
      address = {Porto, Portugal},
      month = {Oct.},
      year = {2012},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ismir12a.pdf}
    }
    
    T. Gerber, M. Dutasta, L. Girin & C. Févotte Professionally-produced music separation guided by covers 2012 Proc.~International Society for Music Information Retrieval Conference (ISMIR)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{ismir12b,
      author = {T.~Gerber and M.~Dutasta and L.~Girin and C.~Févotte} ,
      title = {Professionally-produced music separation guided by covers},
      booktitle = {Proc.~International Society for Music Information Retrieval Conference (ISMIR)},
      address = {Porto, Portugal},
      month = {Oct.},
      year = {2012},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ismir12b.pdf}
    }
    
    B. King, C. Févotte & P. Smaragdis Optimal cost function and magnitude power for NMF-based speech separation and music interpolation 2012 Proc.~IEEE International Workshop on Machine Learning for Signal Processing (MLSP)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{mlsp12,
      author = {B.~King and C.~Févotte and P.~Smaragdis} ,
      title = {Optimal cost function and magnitude power for NMF-based speech separation and music interpolation},
      booktitle = {Proc.~IEEE International Workshop on Machine Learning for Signal Processing (MLSP)},
      address = {Santander, Spain},
      month = {Sep.},
      year = {2012},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/mlsp12.pdf}
    }
    
    J. Hershey, C. Févotte & J. Le~Roux Method for transforming non-stationary signals using a dynamic model 2012   patent [PDF]  
    BibTeX:
    
    @patent{patent2013,
      author = {J.~Hershey and C.~ Févotte and J.~Le~Roux} ,
      title = {Method for transforming non-stationary signals using a dynamic model},
      month = {Oct.},
      year = {2012},
      number = {13657077},
      note = {US Patent 13657077, filed.},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/patents/application14.pdf}
    }
    
    O. Dikmen & C. Févotte Maximum marginal likelihood estimation for nonnegative dictionary learning in the Gamma-Poisson Model 2012 IEEE Transactions on Signal Processing   article [PDF] [DOI] [CODE]  
    Abstract: In this paper we describe an alternative to standard nonnegative matrix factorization (NMF) for nonnegative dictionary learning, i.e., the task of learning a dictionary with nonnegative values from nonnegative data, under the assumption of nonnegative expansion coefficients. A popular cost function used for NMF is the Kullback-Leibler divergence, which underlies a Poisson observation model. NMF can thus be considered as maximization of the joint likelihood of the dictionary and the expansion coefficients. This approach lacks optimality because the number of parameters (which include the expansion coefficients) grows with the number of observations. In this paper we describe variational Bayes and Monte-Carlo EM algorithms for optimization of the marginal likelihood, i.e., the likelihood of the dictionary where the expansion coefficients have been integrated out (given a Gamma prior). We compare the output of both maximum joint likelihood estimation (i.e., standard NMF) and maximum marginal likelihood estimation (MMLE) on real and synthetical datasets. In particular we present face reconstruction results on CBCL dataset and text retrieval results over the musiXmatch dataset, a collection of word counts in song lyrics. The MMLE approach is shown to prevent overfitting by automatically pruning out irrelevant dictionary columns, i.e., embedding automatic model order selection.
    BibTeX:
    
    @article{tspgap,
      author = {O.~Dikmen and C.~Févotte} ,
      title = {Maximum marginal likelihood estimation for nonnegative dictionary learning in the Gamma-Poisson Model},
      journal = {IEEE Transactions on Signal Processing},
      month = {Oct.},
      year = {2012},
      volume = {60},
      number = {10},
      pages = {5163--5175},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_sp_mmle.pdf},
      doi = {http://dx.doi.org/10.1109/TSP.2012.2207117}
    }
    
    C. Févotte & J. Idier Algorithms for nonnegative matrix factorization with the beta-divergence 2011 Neural Computation   article [PDF] [DOI] [CODE]  
    Abstract: This paper describes algorithms for nonnegative matrix factorization (NMF) with the beta-divergence (beta-NMF). The beta-divergence is a family of cost functions parametrized by a single shape parameter beta that takes the Euclidean distance, the Kullback-Leibler divergence and the Itakura-Saito divergence as special cases (beta = 2,1,0, respectively). The proposed algorithms are based on a surrogate auxiliary function (a local majorization of the criterion function). We first describe a majorization-minimization (MM) algorithm that leads to multiplicative updates, which differ from standard heuristic multiplicative updates by a beta-dependent power exponent. The monotonicity of the heuristic algorithm can however be proven for beta in (0,1) using the proposed auxiliary function. Then we introduce the concept of majorization-equalization (ME) algorithm which produces updates that move along constant level sets of the auxiliary function and lead to larger steps than MM. Simulations on synthetic and real data illustrate the faster convergence of the ME approach. The paper also describes how the proposed algorithms can be adapted to two common variants of NMF : penalized NMF (i.e., when a penalty function of the factors is added to the criterion function) and convex-NMF (when the dictionary is assumed to belong to a known subspace).
    BibTeX:
    
    @article{betanmf,
      author = {C.~Févotte and J.~Idier} ,
      title = {Algorithms for nonnegative matrix factorization with the beta-divergence},
      journal = {Neural Computation},
      month = {Sep.},
      year = {2011},
      volume = {23},
      number = {9},
      pages = {2421--2456},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/neco11.pdf},
      doi = {http://dx.doi.org/10.1162/NECO_a_00168}
    }
    
    C. Févotte & J. Idier Algorithmes de factorisation en matrices non-négatives fondée sur la beta-divergence 2011 Proc.~23e colloque GRETSI sur le Traitement du Signal et des Images   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{gretsi11a,
      author = {Févotte, C. and Idier, J.} ,
      title = {Algorithmes de factorisation en matrices non-négatives fondée sur la beta-divergence},
      booktitle = {Proc.~23e colloque GRETSI sur le Traitement du Signal et des Images},
      address = {Bordeaux, France},
      month = {Sep.},
      year = {2011},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/gretsi11a.pdf}
    }
    
    O. Cappé, C. Févotte & D. Rhodes Algorithme EM en ligne simulé pour la factorisation non-négative probabiliste 2011 Proc.~23e colloque GRETSI sur le Traitement du Signal et des Images   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{gretsi11b,
      author = {Cappé, O. and Févotte, C. and Rhodes, D.} ,
      title = {Algorithme EM en ligne simulé pour la factorisation non-négative probabiliste},
      booktitle = {Proc.~23e colloque GRETSI sur le Traitement du Signal et des Images},
      address = {Bordeaux, France},
      month = {Sep.},
      year = {2011},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/gretsi11b.pdf}
    }
    
    A. Lefèvre, F. Bach & C. Févotte Factorisation de matrices structurée en groupes avec la divergence d'Itakura-Saito 2011 Proc.~23e colloque GRETSI sur le Traitement du Signal et des Images   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{gretsi11c,
      author = {Lefèvre, A. and Bach, F. and Févotte, C.} ,
      title = {Factorisation de matrices structurée en groupes avec la divergence d'Itakura-Saito},
      booktitle = {Proc.~23e colloque GRETSI sur le Traitement du Signal et des Images},
      address = {Bordeaux, France},
      month = {Sep.},
      year = {2011},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/gretsi11c.pdf}
    }
    
    C. Févotte Majorization-minimization algorithm for smooth Itakura-Saito nonnegative matrix factorization 2011 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF] [CODE] [DEMO]  
    BibTeX:
    
    @inproceedings{icassp11a,
      author = {C.~Févotte} ,
      title = {Majorization-minimization algorithm for smooth Itakura-Saito nonnegative matrix factorization},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Prague, Czech Republic},
      month = {May},
      year = {2011},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp11a.pdf}
    }
    
    O. Dikmen & C. Févotte Maximum marginal likelihood estimation for nonnegative dictionary learning 2011 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF] [CODE]  
    BibTeX:
    
    @inproceedings{icassp11b,
      author = {O.~Dikmen and C.~Févotte} ,
      title = {Maximum marginal likelihood estimation for nonnegative dictionary learning},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Prague, Czech Republic},
      month = {May},
      year = {2011},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp11b.pdf}
    }
    
    A. Lefèvre, F. Bach & C. Févotte Itakura-Saito nonnegative matrix factorization with group sparsity 2011 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{icassp11c,
      author = {A.~Lefèvre and F.~Bach and C.~Févotte} ,
      title = {Itakura-Saito nonnegative matrix factorization with group sparsity},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Prague, Czech Republic},
      month = {May},
      year = {2011},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp11c.pdf}
    }
    
    A. Ozerov, C. Févotte, R. Blouet & J.-L. Durrieu Multichannel nonnegative tensor factorization with structured constraints for user-guided audio source separation 2011 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{icassp11d,
      author = {A.~Ozerov and C.~Févotte and R.~Blouet and J.-L.~Durrieu} ,
      title = {Multichannel nonnegative tensor factorization with structured constraints for user-guided audio source separation},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Prague, Czech Republic},
      month = {May},
      year = {2011},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp11d.pdf}
    }
    
    L. Oudre, C. Févotte & Y. Grenier Probabilistic template-based chord recognition 2011 IEEE Transactions on Audio, Speech and Language Processing   article [PDF] [DOI] [CODE]  
    Abstract: This paper describes a probabilistic approach to template-based chord recognition in music signals. The algorithm only takes chromagram data and a user-defined dictionary of chord templates as input data. No training or musical information such as key, rhythm or chord transition models is required. The chord occurrences are treated as probabilistic events, whose probabilities are learned from the song using an Expectation-Maximization (EM) algorithm. The adaptative estimation of these probabilities (together with an ad-hoc post-processing filtering) has the desirable effect of smoothing out spurious chords that would occur in our previous baseline work. Our algorithm is compared to various methods that entered the Music Information Retrieval Evaluation eXchange (MIREX) in 2008 and 2009, using a diverse set of evaluation metrics, some of which are new. The systems are tested on two evaluation corpuses; the first one is composed of the Beatles catalog (180 pop-rock songs) and the other one is constituted of 20 songs from various artists and music genres. Results show that our method outperforms state-of-the-art chord recognition systems.
    BibTeX:
    
    @article{ieee_taslp_2011a,
      author = {L.~Oudre and C.~Févotte and Y.~Grenier} ,
      title = {Probabilistic template-based chord recognition},
      journal = {IEEE Transactions on Audio, Speech and Language Processing},
      month = {Nov.},
      year = {2011},
      volume = {19},
      number = {8},
      pages = {2249 -- 2259},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_asl_probachord.pdf},
      doi = {http://dx.doi.org/10.1109/TASL.2010.2098870}
    }
    
    L. Oudre, Y. Grenier & C. Févotte Chord recognition by fitting rescaled chroma vectors to chord templates 2011 IEEE Transactions on Audio, Speech and Language Processing   article [PDF] [DOI] [CODE]  
    BibTeX:
    
    @article{ieee_taslp_2011b,
      author = {L.~Oudre and Y.~Grenier and C.~Févotte} ,
      title = {Chord recognition by fitting rescaled chroma vectors to chord templates},
      journal = {IEEE Transactions on Audio, Speech and Language Processing},
      month = {Sep.},
      year = {2011},
      volume = {19},
      number = {7},
      pages = {2222 -- 2233},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_asl_deterchord.pdf},
      doi = {http://dx.doi.org/10.1109/TASL.2011.2139205}
    }
    
    O. Dikmen & C. Févotte Nonnegative dictionary learning in the exponential noise model for adaptive music signal representation 2011 Advances in Neural Information Processing Systems (NIPS)   inproceedings [PDF] [CODE]  
    Abstract: In this paper we describe a maximum likelihood approach for dictionary learning in the multiplicative exponential noise model. This model is prevalent in audio signal processing where it underlies a generative composite model of the power spectrogram. Maximum joint likelihood estimation of the dictionary and expansion coefficients leads to a nonnegative matrix factorization problem where the Itakura-Saito divergence is used. The optimality of this approach is in question because the number of parameters (which include the expansion coefficients) grows with the number of observations. In this paper we describe a variational procedure for optimization of the marginal likelihood, i.e., the likelihood of the dictionary where the activation coefficients have been integrated out (given a specific prior). We compare the output of both maximum joint likelihood estimation (i.e., standard Itakura-Saito NMF) and maximum marginal likelihood estimation (MMLE) on real and synthetical datasets. The MMLE approach is shown to embed automatic model order selection, akin to automatic relevance determination.
    BibTeX:
    
    @inproceedings{nips11,
      author = {O.~Dikmen and C.~Févotte} ,
      title = {Nonnegative dictionary learning in the exponential noise model for adaptive music signal representation},
      booktitle = {Advances in Neural Information Processing Systems (NIPS)},
      address = {Granada, Spain},
      month = {Dec.},
      year = {2011},
      pages = {2267--2275},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/nips11.pdf}
    }
    
    V.Y.F. Tan & C. Févotte Automatic relevance determination in nonnegative matrix factorization with the beta-divergence 2011 Proc.~NIPS workshop on Sparse Representation and Low-rank Approximation   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{nipswork11,
      author = {V.~Y.~F.~Tan and C.~Févotte} ,
      title = {Automatic relevance determination in nonnegative matrix factorization with the beta-divergence},
      booktitle = {Proc.~NIPS workshop on Sparse Representation and Low-rank Approximation},
      address = {Sierra Nevada, Spain},
      month = {Dec.},
      year = {2011},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/nips11_workshop.pdf}
    }
    
    A. Lefèvre, F. Bach & C. Févotte Online algorithms for nonnegative matrix factorization with the Itakura-Saito divergence 2011 Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{onlineisnmf,
      author = {A.~Lefèvre and F.~Bach and C.~Févotte} ,
      title = {Online algorithms for nonnegative matrix factorization with the Itakura-Saito divergence},
      booktitle = {Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)},
      address = {Mohonk, NY},
      month = {Oct.},
      year = {2011},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/waspaa11.pdf}
    }
    
    A. Ozerov, C. Févotte & R. Blouet Automatic source separation via joint use of segmental information and spatial diversity 2011   patent [PDF]  
    BibTeX:
    
    @patent{patent11,
      author = {A.~Ozerov and C.~Févotte and R.~Blouet} ,
      title = {Automatic source separation via joint use of segmental information and spatial diversity},
      month = {Feb.},
      year = {2011},
      number = {13021692},
      note = {US Patent 13021692, filed.},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/patents/application11.pdf}
    }
    
    C. Févotte, O. Cappé & A.T. Cemgil Efficient Markov chain Monte Carlo inference in composite models with space alternating data augmentation 2011 Proc.~IEEE Workshop on Statistical Signal Processing (SSP)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{ssp11,
      author = {C.~Févotte and O.~Cappé and A.~T.~Cemgil} ,
      title = {Efficient Markov chain Monte Carlo inference in composite models with space alternating data augmentation},
      booktitle = {Proc.~IEEE Workshop on Statistical Signal Processing (SSP)},
      address = {Nice, France},
      month = {June},
      year = {2011},
      pages = {221 -- 224},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ssp11.pdf}
    }
    
    C. Févotte & A. Ozerov Notes on nonnegative tensor factorization of the spectrogram for audio source separation~: statistical insights and towards self-clustering of the spatial cues 2010 Proc.~7th International Symposium on Computer Music Modeling and Retrieval (CMMR)   inproceedings [PDF] [CODE] [DEMO]  
    Abstract: Nonnegative tensor factorization (NTF) of multichannel spectrograms under PARAFAC structure has recently been proposed by Fitzgerald et al as a mean of performing blind source separation (BSS) of multichannel audio data. In this paper we investigate the statistical source models implied by this approach. We show that it implicitly assumes a nonpoint-source model contrasting with usual BSS assumptions and we clarify the links between the measure of fit chosen for the NTF and the implied statistical distribution of the sources. While the original approach of Fitzgeral et al requires a posterior clustering of the spatial cues to group the NTF components into sources, we discuss means of performing the clustering within the factorization. In the results section we test the impact of the simplifying nonpoint-source assumption on underdetermined linear instantaneous mixtures of musical sources and discuss the limits of the approach for such mixtures.
    BibTeX:
    
    @inproceedings{cmmr10,
      author = {C.~Févotte and A.~Ozerov} ,
      title = {Notes on nonnegative tensor factorization of the spectrogram for audio source separation~: statistical insights and towards self-clustering of the spatial cues},
      booktitle = {Proc.~7th International Symposium on Computer Music Modeling and Retrieval (CMMR)},
      editor = {S.~Ystad and M.~ Aramaki and R.~Kronland-Martinet and K.~Jensen},
      publisher = {Springer},
      address = {Málaga, Spain, 2010.},
      month = {June},
      year = {2010},
      volume = {6684},
      pages = {102-115},
      note = {Long paper.},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/cmmr10.pdf}
    }
    
    J.-L. Durrieu, G. Richard, B. David & C. Févotte Source/Filter model for unsupervised main melody extraction from polyphonic audio signals 2010 IEEE Transactions on Audio, Speech and Language Processing   article [PDF] [DOI]  
    Abstract: Extracting the main melody from a polyphonic music recording seems natural even to untrained human listeners. To a certain extent it is related to the concept of source separation, with the human ability of focusing on a specific source in order to extract relevant information. In this paper, we propose a new approach for the estimation and extraction of the main melody (and in particular the leading vocal part) from polyphonic audio signals. To that aim, we propose a new signal model where the leading vocal part is explicitly represented by a specific source/filter model. The proposed representation is investigated in the framework of two statistical models: a Gaussian Scaled Mixture Model (GSMM) and an extended Instantaneous Mixture Model (IMM). For both models, the estimation of the different parameters is done within a maximum-likelihood framework adapted from single-channel source separation techniques. The desired sequence of fundamental frequencies is then inferred from the estimated parameters. The results obtained in a recent evaluation campaign (MIREX08) show that the proposed approaches are very promising and reach state-of-the-art performances on all test sets.
    BibTeX:
    
    @article{durr10,
      author = {J.-L.~Durrieu and G.~Richard and B.~David and C.~Févotte} ,
      title = {Source/Filter model for unsupervised main melody extraction from polyphonic audio signals},
      journal = {IEEE Transactions on Audio, Speech and Language Processing},
      month = {Mar.},
      year = {2010},
      volume = {18},
      number = {3},
      pages = {564--575},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_asl_voice_extrac.pdf},
      doi = {http://dx.doi.org/10.1109/TASL.2010.2041114}
    }
    
    H. Lantéri, C. Theys, C. Richard & C. Févotte Split gradient method for nonnegative matrix factorization 2010 Proc.~18th European Signal Processing Conference (EUSIPCO)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{eusipco10,
      author = {H.~Lantéri and C.~Theys and C.~Richard and C.~Févotte} ,
      title = {Split gradient method for nonnegative matrix factorization},
      booktitle = {Proc.~18th European Signal Processing Conference (EUSIPCO)},
      address = {Aalborg, Denmark},
      month = {Aug.},
      year = {2010},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/eusipco10.pdf}
    }
    
    A. Ozerov & C. Févotte Multichannel nonnegative matrix factorization in convolutive mixtures for audio source separation 2010 IEEE Transactions on Audio, Speech and Language Processing   article [PDF] [DOI] [CODE] [DEMO]  
    Abstract: 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).
    BibTeX:
    
    @article{ieee_asl10,
      author = {A.~Ozerov and C.~Févotte} ,
      title = {Multichannel nonnegative matrix factorization in convolutive mixtures for audio source separation},
      journal = {IEEE Transactions on Audio, Speech and Language Processing},
      month = {Mar.},
      year = {2010},
      volume = {18},
      number = {3},
      pages = {550--563},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_asl_multinmf.pdf},
      doi = {http://dx.doi.org/10.1109/TASL.2009.2031510}
    }
    
    L. Oudre, C. Févotte & Y. Grenier Probabilistic framework for template-based chord recognition 2010 Proc.~IEEE International Workshop on Multimedia Signal Processing (MMSP)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{mmsp10,
      author = {L.~Oudre and C.~Févotte and Y.~Grenier} ,
      title = {Probabilistic framework for template-based chord recognition},
      booktitle = {Proc.~IEEE International Workshop on Multimedia Signal Processing (MMSP)},
      address = {St-Malo, France},
      month = {Oct.},
      year = {2010},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/mmsp10.pdf}
    }
    
    C. Févotte Itakura-Saito nonnegative factorizations of the power spectrogram for music signal decomposition 2010 Machine Audition: Principles, Algorithms and Systems   incollection [PDF] [DOI] [DEMO]  
    Abstract: Nonnegative matrix factorization (NMF) is a popular linear regression technique in the fields of machine learning and signal/image processing. Much research about this topic has been driven by applications in audio. NMF has been for example applied with success to automatic music transcription and audio source separation, where the data is usually taken as the magnitude spectrogram of the sound signal, and the Euclidean distance or Kullback-Leibler divergence are used as measures of fit between the original spectrogram and its approximate factorization. In this chapter we give evidence of the relevance of considering factorization of the power spectrogram, with the Itakura-Saito (IS) divergence. Indeed, IS-NMF is shown to be connected to maximum likelihood inference of variance parameters in a well-defined statistical model of superimposed Gaussian components and this model is in turn shown to be well suited to audio. Furthermore, the statistical setting opens doors to Bayesian approaches and to a variety of computational inference techniques. We discuss in particular model order selection strategies and Markov regularization of the activation matrix, to account for time-persistence in audio. This chapter also discusses extensions of NMF to the multichannel case, in both instantaneous or convolutive recordings, possibly underdetermined. We present in particular audio source separation results of a real stereo musical excerpt.
    BibTeX:
    
    @incollection{wangbook,
      author = {C.~Févotte} ,
      title = {Itakura-Saito nonnegative factorizations of the power spectrogram for music signal decomposition},
      booktitle = {Machine Audition: Principles, Algorithms and Systems},
      editor = {Wenwu Wang},
      publisher = {IGI Global Press},
      month = {Aug.},
      year = {2010},
      chapter = {11},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/chapters/isnmf.pdf},
      doi = {http://dx.doi.org/10.4018/978-1-61520-919-4}
    }
    
    C. Févotte & A.T. Cemgil Nonnegative matrix factorisations as probabilistic inference in composite models 2009 Proc.~17th European Signal Processing Conference (EUSIPCO)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{eusipco09,
      author = {C.~Févotte and A.~T.~Cemgil} ,
      title = {Nonnegative matrix factorisations as probabilistic inference in composite models},
      booktitle = {Proc.~17th European Signal Processing Conference (EUSIPCO)},
      address = {Glasgow, Scotland},
      month = {Aug.},
      year = {2009},
      pages = {1913--1917},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/eusipco09a.pdf}
    }
    
    J.-L. Durrieu, A. Ozerov, C. Févotte, G. Richard & B. David Main instrument separation from stereophonic audio signals using a source/filter models 2009 Proc.~17th European Signal Processing Conference (EUSIPCO)   inproceedings [PDF] [DEMO]  
    BibTeX:
    
    @inproceedings{eusipco09b,
      author = {J.-L.~Durrieu and A.~Ozerov and C.~Févotte and G.~Richard and B.~David} ,
      title = {Main instrument separation from stereophonic audio signals using a source/filter models},
      booktitle = {Proc.~17th European Signal Processing Conference (EUSIPCO)},
      address = {Glasgow, Scotland},
      month = {Aug.},
      year = {2009},
      pages = {15--19},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/eusipco09b.pdf}
    }
    
    A. Ozerov & C. Févotte Multichannel nonnegative matrix factorization in convolutive mixtures. With application to blind audio source separation 2009 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF] [DOI] [DEMO]  
    BibTeX:
    
    @inproceedings{icassp09a,
      author = {A.~Ozerov and C.~Févotte} ,
      title = {Multichannel nonnegative matrix factorization in convolutive mixtures. With application to blind audio source separation},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Taipei, Taiwan},
      month = {Apr.},
      year = {2009},
      pages = {3137-3140},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp09a.pdf},
      doi = {http://dx.doi.org/10.1109/ICASSP.2009.4960289}
    }
    
    N. Bertin, C. Févotte & R. Badeau A tempering approach for Itakura-Saito non-negative matrix factorization. With application to music transcription 2009 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF] [DOI]  
    BibTeX:
    
    @inproceedings{icassp09b,
      author = {N.~Bertin and C.~Févotte and R.~Badeau} ,
      title = {A tempering approach for Itakura-Saito non-negative matrix factorization. With application to music transcription},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Taipei, Taiwan},
      month = {Apr.},
      year = {2009},
      pages = {1545 --1548},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp09b.pdf},
      doi = {http://dx.doi.org/10.1109/ICASSP.2009.4959891}
    }
    
    L. Oudre, Y. Grenier & C. Févotte Template-based chord recognition : influence of the chord types 2009 Proc.~International Society for Music Information Retrieval Conference (ISMIR)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{ismir09,
      author = {L.~Oudre and Y.~Grenier and C.~Févotte} ,
      title = {Template-based chord recognition : influence of the chord types},
      booktitle = {Proc.~International Society for Music Information Retrieval Conference (ISMIR)},
      address = {Kobe, Japan},
      month = {Oct.},
      year = {2009},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ismir09.pdf}
    }
    
    C. Févotte, N. Bertin & J.-L. Durrieu Nonnegative matrix factorization with the Itakura-Saito divergence. With application to music analysis 2009 Neural Computation   article [PDF] [DOI] [CODE] [DEMO]  
    Abstract: This letter presents theoretical, algorithmic, and experimental results about nonnegative matrix factorization (NMF) with the Itakura-Saito (IS) divergence. We describe how IS-NMF is underlaid by a well-defined statistical model of superimposed gaussian components and is equivalent to maximum likelihood estimation of variance parameters. This setting can accommodate regularization constraints on the factors through Bayesian priors. In particular, inverse-gamma and gamma Markov chain priors are considered in this work. Estimation can be carried out using a space-alternating generalized expectation-maximization (SAGE) algorithm; this leads to a novel type of NMF algorithm, whose convergence to a stationary point of the IS cost function is guaranteed.

    We also discuss the links between the IS divergence and other cost functions used in NMF, in particular, the Euclidean distance and the generalized Kullback-Leibler (KL) divergence. As such, we describe how IS-NMF can also be performed using a gradient multiplicative algorithm (a standard algorithm structure in NMF) whose convergence is observed in practice, though not proven.

    Finally, we report a furnished experimental comparative study of Euclidean-NMF, KL-NMF, and IS-NMF algorithms applied to the power spectrogram of a short piano sequence recorded in real conditions, with various initializations and model orders. Then we show how IS-NMF can successfully be employed for denoising and upmix (mono to stereo conversion) of an original piece of early jazz music. These experiments indicate that IS-NMF correctly captures the semantics of audio and is better suited to the representation of music signals than NMF with the usual Euclidean and KL costs.

    BibTeX:
    
    @article{neco09,
      author = {C.~Févotte and N.~Bertin and J.-L.~Durrieu} ,
      title = {Nonnegative matrix factorization with the Itakura-Saito divergence. With application to music analysis},
      journal = {Neural Computation},
      month = {Mar.},
      year = {2009},
      volume = {21},
      number = {3},
      pages = {793-830},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/neco09_is-nmf.pdf},
      doi = {http://dx.doi.org/10.1162/neco.2008.04-08-771}
    }
    
    V.Y.F. Tan & C. Févotte Automatic relevance determination in nonnegative matrix factorization 2009 Proc.~Workshop on Signal Processing with Adaptative Sparse Structured Representations (SPARS)   inproceedings [PDF] [CODE]  
    BibTeX:
    
    @inproceedings{spars09,
      author = {V.~Y.~F.~Tan and C.~Févotte} ,
      title = {Automatic relevance determination in nonnegative matrix factorization},
      booktitle = {Proc.~Workshop on Signal Processing with Adaptative Sparse Structured Representations (SPARS)},
      address = {St-Malo, France},
      month = {Apr.},
      year = {2009},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/spars09.pdf}
    }
    
    A. Ozerov, C. Févotte & M. Charbit Factorial scaled hidden Markov model for polyphonic audio representation and source separation 2009 Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)   inproceedings [PDF] [DEMO]  
    BibTeX:
    
    @inproceedings{waspaa09-1,
      author = {A.~Ozerov and C.~Févotte and M.~Charbit} ,
      title = {Factorial scaled hidden Markov model for polyphonic audio representation and source separation},
      booktitle = {Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)},
      address = {Mohonk, NY, USA},
      month = {Oct.},
      year = {2009},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/waspaa09a.pdf}
    }
    
    L. Oudre, Y. Grenier & C. Févotte Chord recognition using measures of fit, chord templates and filtering methods 2009 Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{waspaa09b,
      author = {L.~Oudre and Y.~Grenier and C.~Févotte} ,
      title = {Chord recognition using measures of fit, chord templates and filtering methods},
      booktitle = {Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)},
      address = {Mohonk, NY},
      month = {Oct.},
      year = {2009},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/waspaa09b.pdf}
    }
    
    R. Blouet, G. Rapaport, I. Cohen & C. Févotte Evaluation of several strategies for single sensor speech/music separation 2008 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF] [DOI]  
    BibTeX:
    
    @inproceedings{icassp08,
      author = {R.~Blouet and G.~Rapaport and I.~Cohen and C.~Févotte} ,
      title = {Evaluation of several strategies for single sensor speech/music separation},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Las Vegas, USA},
      month = {Apr.},
      year = {2008},
      pages = {37-40},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp08.pdf},
      doi = {http://dx.doi.org/10.1109/ICASSP.2008.4517540}
    }
    
    C. Févotte, B. Torrésani, L. Daudet & S.J. Godsill Sparse linear regression with structured priors and application to denoising of musical audio 2008 IEEE Transactions on Audio, Speech and Language Processing   article [PDF] [DOI] [CODE] [DEMO]  
    Abstract: We describe in this paper an audio denoising technique based on sparse linear regression with structured priors. The noisy signal is decomposed as a linear combination of atoms belonging to two modified discrete cosine transform (MDCT) bases, plus a residual part containing the noise. One MDCT basis has a long time resolution, and thus high frequency resolution, and is aimed at modeling tonal parts of the signal, while the other MDCT basis has short time resolution and is aimed at modeling transient parts (such as attacks of notes). The problem is formulated within a Bayesian setting. Conditional upon an indicator variable which is either 0 or 1, one expansion coefficient is set to zero or given a hierarchical prior. Structured priors are employed for the indicator variables; using two types of Markov chains, persistency along the time axis is favored for expansion coefficients of the tonal layer, while persistency along the frequency axis is favored for the expansion coefficients of the transient layer. Inference about the denoised signal and model parameters is performed using a Gibbs sampler, a standard Markov chain Monte Carlo (MCMC) sampling technique. We present results for denoising of a short glockenspiel excerpt and a long polyphonic music excerpt. Our approach is compared with unstructured sparse regression and with structured sparse regression in a single resolution MDCT basis (no transient layer). The results show that better denoising is obtained, both from signal-to-noise ratio measurements and from subjective criteria, when both a transient and tonal layer are used, in conjunction with our proposed structured prior framework.
    BibTeX:
    
    @article{ieee_asl07b,
      author = {C.~Févotte and B.~Torrésani and L.~Daudet and S.~J.~Godsill} ,
      title = {Sparse linear regression with structured priors and application to denoising of musical audio},
      journal = {IEEE Transactions on Audio, Speech and Language Processing},
      month = {Jan.},
      year = {2008},
      volume = {16},
      number = {1},
      pages = {174--185},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_asl_sparsereg_struc.pdf},
      doi = {http://dx.doi.org/10.1109/TASL.2007.909290}
    }
    
    P. Aimé & C. Févotte La simplification administrative de la gestion des unités de recherche 2008 Rapport de l'Inspection Générale de l'Éducation Nationale et de la Recherche (IGAENR), no 2008-089   misc [PDF]  
    BibTeX:
    
    @misc{igaenr09,
      author = {P. Aimé and C. Févotte} ,
      title = {La simplification administrative de la gestion des unités de recherche},
      month = {Oct.},
      year = {2008},
      note = {http://media.enseignementsup-recherche.gouv.fr/file/Concours_2008/22/3/2008-089simplification_44223.pdf},
      url = {http://media.enseignementsup-recherche.gouv.fr/file/Concours_2008/22/3/2008-089simplification_44223.pdf}
    }
    
    A.T. Cemgil, C. Févotte & S.J. Godsill Variational and stochastic inference for Bayesian Source Separation 2007 Digital Signal Processing   article [PDF]  
    Abstract: We tackle the general linear instantaneous model (possibly underdetermined and noisy) where we model the source prior with a Student t distribution. The conjugate-exponential characterisation of the t distribution as an infinite mixture of scaled Gaussians enables us to do efficient inference.We study two well-known inference methods, Gibbs sampler and variational Bayes for Bayesian source separation. We derive both techniques as local message passing algorithms to highlight their algorithmic similarities and to contrast their different convergence characteristics and computational requirements. Our simulation results suggest that typical posterior distributions in source separation have multiple local maxima. Therefore we propose a hybrid approach where we explore the state space with a Gibbs sampler and then switch to a deterministic algorithm. This approach seems to be able to combine the speed of the variational approach with the robustness of the Gibbs sampler.
    BibTeX:
    
    @article{dsp07,
      author = {A.~T.~Cemgil and C.~Févotte and S.~J.~Godsill} ,
      title = {Variational and stochastic inference for Bayesian Source Separation},
      journal = {Digital Signal Processing},
      month = {Sep.},
      year = {2007},
      volume = {17},
      number = {5},
      pages = {891--913},
      note = {Special issue Bayesian source separation, ed.~E.~E.~Kuruoglu and K.~H.~Knuth},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/elsevier_dsp_variational.pdf}
    }
    
    S.J. Godsill, A.T. Cemgil, C. Févotte & P.J. Wolfe Bayesian computational methods for sparse audio and music processing 2007 Proc.~15th European Signal Processing Conference (EUSIPCO)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{eusipco07,
      author = {S.~J.~Godsill and A.~T.~Cemgil and C.~Févotte and P.~J.~Wolfe} ,
      title = {Bayesian computational methods for sparse audio and music processing},
      booktitle = {Proc.~15th European Signal Processing Conference (EUSIPCO)},
      address = {Poznań, Poland},
      month = {Sep.},
      year = {2007},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/eusipco07.pdf}
    }
    
    C. Févotte & F. Theis Pivot selection strategies in Jacobi joint block-diagonalization 2007 Proc.~7th International Conference on Independent Component Analysis and Signal Separation (ICA)   inproceedings [PDF] [CODE]  
    BibTeX:
    
    @inproceedings{ica07,
      author = {C.~Févotte and F.~Theis} ,
      title = {Pivot selection strategies in Jacobi joint block-diagonalization},
      booktitle = {Proc.~7th International Conference on Independent Component Analysis and Signal Separation (ICA)},
      address = {London, UK},
      month = {Sep.},
      year = {2007},
      pages = {177--187},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ica07.pdf}
    }
    
    C. Févotte Bayesian audio source separation 2007 Blind speech separation   incollection [PDF] [DOI] [DEMO]  
    Abstract: In this chapter we describe a Bayesian approach to audio source separation. The approach relies on probabilistic modeling of sound sources as (sparse) linear combinations of atoms from
    a dictionary and Markov chain Monte Carlo (MCMC) inference. Several prior distributions are considered for the source expansion coefficients. We first consider independent and iden-
    tically distributed (iid) general priors with two choices of distributions. The first one is the Student t, which is a good model for sparsity when the shape parameter has a low value. The
    second one is a hierarchical mixture distribution; conditionally upon an indicator variable, one coefficient is either set to zero or given a normal distribution, whose variance is in turn
    given an inverted-Gamma distribution. Then, we consider more audio-specific models where both the identically distributed and independently distributed assumptions are lifted. Using a
    Modified Discrete Cosine Transform (MDCT) dictionary, a time-frequency orthonormal basis, we describe frequency-dependent structured priors which explicitly model the harmonic struc-
    ture of sound, using a Markov hierarchical modeling of the expansion coefficients. Separation results are given for a stereophonic recording of 3 sources.
    BibTeX:
    
    @incollection{makinobook,
      author = {C.~Févotte} ,
      title = {Bayesian audio source separation},
      booktitle = {Blind speech separation},
      editor = {S.~Makino and T.-W.~Lee and H.~Sawada},
      publisher = {Springer},
      month = {Sep.},
      year = {2007},
      chapter = {11},
      pages = {305--335},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/chapters/bass.pdf},
      doi = {http://www.springer.com/engineering/signals/book/978-1-4020-6478-4}
    }
    
    C. Févotte & F. Theis Orthonormal approximate joint block-diagonalization 2007 GET/Télécom Paris   techreport [PDF] [CODE]  
    BibTeX:
    
    @techreport{simax06,
      author = {C.~Févotte and F.~Theis} ,
      title = {Orthonormal approximate joint block-diagonalization},
      institution = {GET/Télécom Paris},
      month = {Apr.},
      year = {2007},
      number = {2007D007},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/techreps/techrep07.pdf}
    }
    
    C. Févotte Bayesian blind separation of audio mixtures with structured priors 2006 Proc.~14th European Signal Processing Conference (EUSIPCO)   inproceedings [PDF] [DEMO]  
    BibTeX:
    
    @inproceedings{eusipco06,
      author = {C.~Févotte} ,
      title = {Bayesian blind separation of audio mixtures with structured priors},
      booktitle = {Proc.~14th European Signal Processing Conference (EUSIPCO)},
      address = {Florence, Italy},
      month = {Sep.},
      year = {2006},
      note = {Special session Undetermined sparse audio source separation (invited paper)},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/eusipco06.pdf}
    }
    
    C. Févotte & S.J. Godsill Blind separation of sparse sources using Jeffrey's inverse prior and the EM algorithm 2006 Proc.~6th International Conference on Independent Component Analysis and Blind Source Separation (ICA)   inproceedings [PDF] [DEMO]  
    BibTeX:
    
    @inproceedings{ica06,
      author = {C.~Févotte and S.~J.~Godsill} ,
      title = {Blind separation of sparse sources using Jeffrey's inverse prior and the EM algorithm},
      booktitle = {Proc.~6th International Conference on Independent Component Analysis and Blind Source Separation (ICA)},
      address = {Charleston, SC, USA},
      month = {Mar.},
      year = {2006},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ica06.pdf}
    }
    
    F. Desobry & C. Févotte Kernel PCA based estimation of the mixing matrix in linear instantaneous mixtures of sparse sources 2006 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF] [DOI]  
    BibTeX:
    
    @inproceedings{icassp06-1,
      author = {F.~Desobry and C.~Févotte} ,
      title = {Kernel PCA based estimation of the mixing matrix in linear instantaneous mixtures of sparse sources},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Toulouse, France},
      month = {May},
      year = {2006},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp06a.pdf},
      doi = {http://dx.doi.org/10.1109/ICASSP.2006.1661364}
    }
    
    C. Févotte, L. Daudet, S.J. Godsill & B. Torrésani Sparse regression with structured priors: application to audio denoising 2006 Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)   inproceedings [PDF] [DOI] [DEMO]  
    BibTeX:
    
    @inproceedings{icassp06-2,
      author = {C.~Févotte and L.~Daudet and S.~J.~Godsill and B.~Torrésani} ,
      title = {Sparse regression with structured priors: application to audio denoising},
      booktitle = {Proc.~IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
      address = {Toulouse, France},
      month = {May},
      year = {2006},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/icassp06b.pdf},
      doi = {http://dx.doi.org/10.1109/ICASSP.2006.1660589}
    }
    
    E. Vincent, R. Gribonval & C. Févotte Performance measurement in blind audio source separation 2006 IEEE Transactions on Audio, Speech and Language Processing   article [PDF] [DOI] [CODE] [DEMO]  
    Abstract: In this paper, we discuss the evaluation of blind audio source separation (BASS) algorithms. Depending on the exact application, different distortions can be allowed between an estimated source and the wanted true source. We consider four different sets of such allowed distortions, from time-invariant gains to time-varying filters. In each case, we decompose the estimated source into a true source part plus error terms corresponding to interferences, additive noise, and algorithmic artifacts. Then, we derive a global performance measure using an energy ratio, plus a separate performance measure for each error term. These measures are computed and discussed on the results of several BASS problems with various difficulty levels
    BibTeX:
    
    @article{ieee_asl06,
      author = {E.~Vincent and R.~Gribonval and C.~Févotte} ,
      title = {Performance measurement in blind audio source separation},
      journal = {IEEE Transactions on Audio, Speech and Language Processing},
      month = {Jul.},
      year = {2006},
      volume = {14},
      number = {4},
      pages = {1462--1469},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_asl_bsseval.pdf},
      doi = {http://dx.doi.org/10.1109/TSA.2005.858005}
    }
    
    C. Févotte & S.J. Godsill A Bayesian approach to blind separation of sparse sources 2006 IEEE Transactions on Audio, Speech and Language Processing   article [PDF] [DOI] [DEMO]  
    Abstract: In this paper, we discuss the evaluation of blind audio source separation (BASS) algorithms. Depending on the exact application, different distortions can be allowed between an estimated source and the wanted true source. We consider four different sets of such allowed distortions, from time-invariant gains to time-varying filters. In each case, we decompose the estimated source into a true source part plus error terms corresponding to interferences, additive noise, and algorithmic artifacts. Then, we derive a global performance measure using an energy ratio, plus a separate performance measure for each error term. These measures are computed and discussed on the results of several BASS problems with various difficulty levels
    BibTeX:
    
    @article{ieee_asl07,
      author = {C.~Févotte and S.~J.~Godsill} ,
      title = {A Bayesian approach to blind separation of sparse sources},
      journal = {IEEE Transactions on Audio, Speech and Language Processing},
      month = {Nov.},
      year = {2006},
      volume = {14},
      number = {6},
      pages = {2174--2188},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_asl_sparsebss.pdf},
      doi = {http://dx.doi.org/10.1109/TSA.2005.858523}
    }
    
    C. Févotte & S.J. Godsill Sparse linear regression in unions of bases via Bayesian variable selection 2006 IEEE Signal Processing Letters   article [PDF] [DOI] [DEMO]  
    Abstract: In this letter, we propose an approach for sparse linear regression in unions of bases inspired by Bayesian variable selection. Conditionally upon an indicator variable that is 0 or 1, one expansion coefficient of the signal corresponding to one atom of the dictionary is either set to zero or given a Student t prior. A Gibbs sampler (a standard Markov chain Monte Carlo technique) is used to sample from the posterior distribution of the indicator variables, the expansion coefficients (corresponding to nonzero indicator variables), the hyperparameters of the Student t priors, and the variance of the residual signal. The structure of the dictionary, assumed to be a union of bases, allows for alternate sampling of the indicator variables and the expansion coefficients from each basis and avoids any large matrix inversion. Our method is applied to the denoising problem of a piano sequence, using a dual-resolution union of two modified discrete cosine transform bases
    BibTeX:
    
    @article{ieee_spl06,
      author = {C.~Févotte and S.~J.~Godsill} ,
      title = {Sparse linear regression in unions of bases via Bayesian variable selection},
      journal = {IEEE Signal Processing Letters},
      month = {Jul.},
      year = {2006},
      volume = {13},
      number = {7},
      pages = {441--444},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_spl_sparsereg.pdf},
      doi = {http://dx.doi.org/10.1109/LSP.2006.873139}
    }
    
    L. Benaroya, R. Blouet, C Févotte & I. Cohen Single sensor source separation using multiple-window STFT representation 2006 Proc. International Workshop on Acoustic Echo and Noise Control (IWAENC)   inproceedings [PDF] [DEMO]  
    BibTeX:
    
    @inproceedings{iwaenc06,
      author = {L.~Benaroya and R.~Blouet and C~Févotte and I.~Cohen} ,
      title = {Single sensor source separation using multiple-window STFT representation},
      booktitle = {Proc. International Workshop on Acoustic Echo and Noise Control (IWAENC)},
      address = {Paris, France},
      month = {Sep.},
      year = {2006},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/iwaenc06.pdf}
    }
    
    A.T. Cemgil, C. Févotte & S.J. Godsill Blind separation of sparse sources using variational EM 2005 Proc.~13th European Signal Processing Conference (EUSIPCO)   inproceedings [PDF] [DEMO]  
    BibTeX:
    
    @inproceedings{eusipco05,
      author = {A.~T.~Cemgil and C.~Févotte and S.~J.~Godsill} ,
      title = {Blind separation of sparse sources using variational EM},
      booktitle = {Proc.~13th European Signal Processing Conference (EUSIPCO)},
      address = {Antalya, Turkey},
      month = {Sep.},
      year = {2005},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/eusipco05.pdf}
    }
    
    V.Y.F. Tan & C. Févotte A study of the effect of source sparsity for various transforms on blind audio source separation performance 2005 Proc.~Workshop on Signal Processing with Adaptative Sparse Structured Representations (SPARS)   inproceedings [PDF] [DEMO]  
    BibTeX:
    
    @inproceedings{spars05,
      author = {V.~Y.~F.~Tan and C.~Févotte} ,
      title = {A study of the effect of source sparsity for various transforms on blind audio source separation performance},
      booktitle = {Proc.~Workshop on Signal Processing with Adaptative Sparse Structured Representations (SPARS)},
      address = {Rennes, France},
      month = {Nov.},
      year = {2005},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/spars05.pdf}
    }
    
    C. Févotte, R. Gribonval & E. Vincent BSS-EVAL Toolbox User Guide~: Revision 2.0 2005 IRISA   techreport [PDF]  
    BibTeX:
    
    @techreport{tr05,
      author = {C.~Févotte and R.~Gribonval and E.~Vincent} ,
      title = {BSS-EVAL Toolbox User Guide~: Revision 2.0},
      institution = {IRISA},
      month = {Apr.},
      year = {2005},
      number = {1706},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/techreps/BSSEVAL2userguide.pdf}
    }
    
    C. Févotte & J.-F. Cardoso Maximum likelihood approach for blind audio source separation using time-frequency Gaussian models 2005 Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)   inproceedings [PDF] [DOI] [DEMO]  
    BibTeX:
    
    @inproceedings{waspaa05-1,
      author = {C.~Févotte and J.-F.~Cardoso} ,
      title = {Maximum likelihood approach for blind audio source separation using time-frequency Gaussian models},
      booktitle = {Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)},
      address = {Mohonk, NY, USA},
      month = {Oct.},
      year = {2005},
      pages = {78-81},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/waspaa05a.pdf},
      doi = {http://dx.doi.org/10.1109/ASPAA.2005.1540173}
    }
    
    C. Févotte & S.J. Godsill A Bayesian approach to time-frequency based blind source separation 2005 Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)   inproceedings [PDF] [DEMO]  
    BibTeX:
    
    @inproceedings{waspaa05-2,
      author = {C.~Févotte and S.~J.~Godsill} ,
      title = {A Bayesian approach to time-frequency based blind source separation},
      booktitle = {Proc.~IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)},
      address = {Mohonk, NY, USA},
      month = {Oct.},
      year = {2005},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/waspaa05b.pdf}
    }
    
    C. Févotte, S.J. Godsill & P.J. Wolfe Bayesian approach for blind separation of underdetermined mixtures of sparse sources 2004 Proc. 5th International Conference on Independent Component Analysis and Blind Source Separation (ICA)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{ica04,
      author = {C.~Févotte and S.~J.~Godsill and P.~J.~Wolfe} ,
      title = {Bayesian approach for blind separation of underdetermined mixtures of sparse sources},
      booktitle = {Proc. 5th International Conference on Independent Component Analysis and Blind Source Separation (ICA)},
      address = {Granada, Spain},
      month = {Sep.},
      year = {2004},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ica04.pdf}
    }
    
    D. Farina, C. Févotte, C. Doncarli & R. Merletti Blind separation of linear instantaneous mixtures of non-stationary surface myoelectric signals 2004 IEEE Transactions on Biomedical Engineering   article [PDF] [DOI]  
    Abstract: Electromyographic (EMG) recordings detected over the skin may be mixtures of signals generated by different active muscles due to the phenomena related to volume conduction. Separation of the sources is necessary when single muscle activity has to be detected. Signals generated by different muscles may be considered uncorrelated but in general overlap in time and frequency. Under certain assumptions, mixtures of surface EMG signals can be considered as linear instantaneous but no a priori information about the mixing matrix is available when different muscles are active. In this study, we applied blind source separation (BSS) methods to separate the signals generated by two active muscles during a force-varying task. As the signals are non stationary, an algorithm based on spatial time-frequency distributions was applied on simulated and experimental EMG signals. The experimental signals were collected from the flexor carpi radialis and the pronator teres muscles which could be activated selectively for wrist flexion and rotation, respectively. From the simulations, correlation coefficients between the reference and reconstructed sources were higher than 0.85 for signals largely overlapping both in time and frequency and for signal-to-noise ratios as low as 5 dB. The Choi-Williams and Bessel kernels, in this case, performed better than the Wigner-Ville one. Moreover, the selection of time-frequency points for the procedure of joint diagonalization used in the BSS algorithm significantly influenced the results. For the experimental signals, the interference of the other source in each reconstructed source was significantly attenuated by the application of the BSS method. The ratio between root-mean-square values of the signals from the two sources detected over one of the muscles increased from (mean +/- standard deviation) 2.33 +/- 1.04 to 4.51 +/- 1.37 and from 1.55 +/- 0.46 to 2.72 +/- 0.65 for wrist flexion and rotation, respectively. This increment was statistically significant. It was concluded that the BSS approach applied is promising for the separation of surface EMG signals, with applications ranging from muscle assessment to detection of muscle activation intervals, and to the control of myoelectric prostheses.
    BibTeX:
    
    @article{ieee_biomed04,
      author = {D.~Farina and C.~Févotte and C.~Doncarli and R.~Merletti} ,
      title = {Blind separation of linear instantaneous mixtures of non-stationary surface myoelectric signals},
      journal = {IEEE Transactions on Biomedical Engineering},
      month = {Sep.},
      year = {2004},
      volume = {51},
      number = {9},
      pages = {1555 --1567},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_be_surfaceemgs.pdf},
      doi = {http://dx.doi.org/10.1109/TBME.2004.828048}
    }
    
    C. Févotte & C. Doncarli Two contributions to blind source separation using time-frequency distributions 2004 IEEE Signal Processing Letters   article [PDF] [DOI] [CODE]  
    Abstract: We present two improvements/extensions of a previous deterministic blind source separation (BSS) technique, by Belouchrani and Amin, that involves joint-diagonalization of a set of Cohen's class spatial time-frequency distributions. The first contribution concerns the extension of the BSS technique to the stochastic case using spatial Wigner-Ville spectrum. Then, we show that Belouchrani and Amin's technique can be interpreted as a practical implementation of the general equations we provide in the stochastic case. The second contribution is a new criterion aimed at selecting more efficiently the time-frequency locations where the spatial matrices should be joint-diagonalized, introducing single autoterms selection. Simulation results on stochastic time-varying autoregressive moving average (TVARMA) signals demonstrate the improved efficiency of the method.
    BibTeX:
    
    @article{ieee_spl04,
      author = {C.~Févotte and C.~Doncarli} ,
      title = {Two contributions to blind source separation using time-frequency distributions},
      journal = {IEEE Signal Processing Letters},
      month = {Mar.},
      year = {2004},
      volume = {11},
      number = {3},
      pages = {386-389},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/journals/ieee_spl_tfbss.pdf},
      doi = {http://dx.doi.org/10.1109/LSP.2003.819343}
    }
    
    D. Farina, F. Lebrun, C. Févotte, C. Doncarli & R. Merletti Blind source separation of linear mixtures of non-stationary surface EMG signals 2003 Proc.~19e colloque GRETSI sur le Traitement du Signal et des Images   inproceedings  
    BibTeX:
    
    @inproceedings{gretsi03-1,
      author = {D.~Farina and F.~Lebrun and C.~Févotte and C.~Doncarli and R.~Merletti} ,
      title = {Blind source separation of linear mixtures of non-stationary surface EMG signals},
      booktitle = {Proc.~19e colloque GRETSI sur le Traitement du Signal et des Images},
      address = {Paris, France},
      month = {Sep.},
      year = {2003}
    }
    
    E. Vincent, C. Févotte & R. Gribonval Comment évaluer les algorithmes de séparation audio~? 2003 Proc.~19e colloque GRETSI sur le Traitement du Signal et des Images   inproceedings  
    BibTeX:
    
    @inproceedings{gretsi03-2,
      author = {E.~Vincent and C.~Févotte and R.~Gribonval} ,
      title = {Comment évaluer les algorithmes de séparation audio~?},
      booktitle = {Proc.~19e colloque GRETSI sur le Traitement du Signal et des Images},
      address = {Paris, France},
      month = {Sep.},
      year = {2003}
    }
    
    C. Févotte & C. Doncarli A unified presentation of blind source separation methods for convolutive mixtures using block-diagonalization 2003 Proc.~4th Symposium on Independent Component Analysis and Blind Source Separation (ICA)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{ica03-1,
      author = {C.~Févotte and C.~Doncarli} ,
      title = {A unified presentation of blind source separation methods for convolutive mixtures using block-diagonalization},
      booktitle = {Proc.~4th Symposium on Independent Component Analysis and Blind Source Separation (ICA)},
      address = {Nara, Japan},
      month = {Apr.},
      year = {2003},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ica03a.pdf}
    }
    
    E. Vincent, C. Févotte, R. Gribonval & al A tentative typology of audio source separation tasks 2003 Proc.~4th Symposium on Independent Component Analysis and Blind Source Separation (ICA)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{ica03-2,
      author = {E.~Vincent and C.~Févotte and R.~Gribonval and al} ,
      title = {A tentative typology of audio source separation tasks},
      booktitle = {Proc.~4th Symposium on Independent Component Analysis and Blind Source Separation (ICA)},
      address = {Nara, Japan},
      month = {Apr.},
      year = {2003},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ica03b.pdf}
    }
    
    R. Gribonval, L. Benaroya, E. Vincent & C. Févotte Proposals for performance measurement in source separation 2003 Proc.~4th Symposium on Independent Component Analysis and Blind Source Separation (ICA)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{ica03-3,
      author = {R.~Gribonval and L.~Benaroya and E.~Vincent and C.~Févotte} ,
      title = {Proposals for performance measurement in source separation},
      booktitle = {Proc.~4th Symposium on Independent Component Analysis and Blind Source Separation (ICA)},
      address = {Nara, Japan},
      month = {Apr.},
      year = {2003},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/ica03c.pdf}
    }
    
    C. Févotte, A. Debiolles & C. Doncarli Blind separation of FIR convolutive mixtures: application to speech signals 2003 Proc.~1st ISCA Workshop on Non-Linear Speech Processing   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{nolisp03,
      author = {C.~Févotte and A.~Debiolles and C.~Doncarli} ,
      title = {Blind separation of FIR convolutive mixtures: application to speech signals},
      booktitle = {Proc.~1st ISCA Workshop on Non-Linear Speech Processing},
      address = {Le Croisic, France},
      month = {May},
      year = {2003},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/nolisp03.pdf}
    }
    
    C. Févotte Approche temps-fréquence pour la séparation aveugle de sources non-stationnaires (A time-frequency approach for blind separation of non-stationary sources) 2003 École Centrale de Nantes et Université de Nantes   phdthesis [PDF]  
    BibTeX:
    
    @phdthesis{these,
      author = {C.~Févotte} ,
      title = {Approche temps-fréquence pour la séparation aveugle de sources non-stationnaires (A time-frequency approach for blind separation of non-stationary sources)},
      school = {École Centrale de Nantes et Université de Nantes},
      month = {Oct.},
      year = {2003},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/theses/these_fevotte.pdf}
    }
    
    A. Holobar, C. Févotte, C. Doncarli & D. Zazula Single autoterms selection for blind source separation in time-frequenccy plane 2002 Proc.~11th European Signal Processing Conference (EUSIPCO)   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{eusipco02,
      author = {A.~Holobar and C.~Févotte and C.~Doncarli and D.~Zazula} ,
      title = {Single autoterms selection for blind source separation in time-frequenccy plane},
      booktitle = {Proc.~11th European Signal Processing Conference (EUSIPCO)},
      address = {Toulouse, France},
      month = {Sep.},
      year = {2002},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/eusipco02.pdf}
    }
    
    L. De Lathauwer, C. Févotte, B. De Moor & J. Vandewalle Jacobi algorithm for joint block diagonalization in blind identification 2002 Proc.~23rd Symposium on Information Theory in the Benelux   inproceedings  
    BibTeX:
    
    @inproceedings{info02,
      author = {L.~De~Lathauwer and C.~Févotte and B.~De~Moor and J.~Vandewalle} ,
      title = {Jacobi algorithm for joint block diagonalization in blind identification},
      booktitle = {Proc.~23rd Symposium on Information Theory in the Benelux},
      address = {Louvain-la-Neuve, Belgium},
      month = {May},
      year = {2002},
      pages = {155--162}
    }
    
    E. Le~Carpentier & C. Févotte Séparation de sources autorégressives gaussiennes par maximum de vraisemblance et filtrage de Kalman 2001 Proc.~18e colloque GRETSI sur le Traitement du Signal et des Images   inproceedings [PDF]  
    BibTeX:
    
    @inproceedings{gretsi01,
      author = {E.~Le~Carpentier and C.~Févotte} ,
      title = {Séparation de sources autorégressives gaussiennes par maximum de vraisemblance et filtrage de Kalman},
      booktitle = {Proc.~18e colloque GRETSI sur le Traitement du Signal et des Images},
      address = {Toulouse, France},
      month = {Sep.},
      year = {2001},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/proceedings/gretsi01.pdf}
    }
    
    C. Févotte Acoustique des salles: modélisation de l'environnement sonore (Room acoustics: identification of room transfer functions) 2000 École Centrale de Nantes   mastersthesis [PDF]  
    BibTeX:
    
    @mastersthesis{dea2,
      author = {C.~Févotte} ,
      title = {Acoustique des salles: modélisation de l'environnement sonore (Room acoustics: identification of room transfer functions)},
      school = {École Centrale de Nantes},
      year = {2000},
      url = {https://www.irit.fr/~Cedric.Fevotte/publications/theses/dea_fevotte.pdf}
    }
    

    Created by JabRef on 07/11/2016.