Institut de Mathématiques de Marseille, UMR 7373

Aller au menu
Aller à la recherche
Aller au contenu
Aller à la navigation

Le CNRS
Autres sites CNRS

L’Institut
Annuaire
Équipes de recherche
Séminaires
Colloques
Publications
Contrats de recherche / Projets
Relations internationales
Enseignement-Formation
Bibliothèques
Ressources / Procédures
Offres d’emploi
Appels à projets
- AAP 2017
- AAP 2018
Liens
Trombinoscope
Intranet (AIGLe)
--Revue de Presse CNRS
---Nouveaux Personnels

Rechercher

Accueil >

TORRESANI Bruno

Fonction : Professeur

Organisme : Aix Marseille Université
Professeur
(HDR)

Equipe: Signal et Image (SI)

Bruno.Torresani _AT_

univ-amu.fr

0413551412 (ou 0413551356)

Bureau: 119, Etg: 1, Bât: CMI - Site : Chateau-Gombert

Page Web perso

Administration Nationale:

Expert ANR

Curriculum Vitae:

1988-1998: CR CNRS
1998-: PR, Université de Provence puis Aix-Marseille Université

Domaines de Recherche:

Dernieres productions scientifiques :

Fenêtre et grille optimales pour la transformée de Gabor Exemples d'application à l'analyse audio

Auteur(s): Lachambre H, Ricaud Benjamin, Stempfel G, Torresani B., Wiesmeyr C, Onchis-Moaca Darian

Conference: XXV-ième Colloque GRETSI (Lyon, FR, 2015-09-07)
Actes de conférence: Colloque GRETSI 2015, vol. p. (2015)

Ref HAL: hal-01279441_v1
Exporter : BibTex | endNote
Résumé:

Cet article présente l'utilisation d'une grille optimale et d'une fenêtre optimale pour le calcul de la transformée de Gabor discrète. Dans le cas d'une Gaussienne généralisée, nous étendons des travaux précédents et proposons une fenêtre localement optimale pour des si-gnaux non-stationnaires. Nous présentons des résultats sur trois problèmes d'analyse temps-fréquence, sur des signaux réels et synthétiques : la distinction de composantes temps-fréquence proches, l'estimation de fréquence instantané et l'estimation du Rapport Signal à Bruit. Abstract – This article deals with the use of optimal lattice and optimal window in Discrete Gabor Transform computation. In the case of a generalized Gaussian window, extending earlier contributions, we introduce an additional local window adaptation technique for non-stationary signals. We illustrate our approach and the earlier one by addressing three time-frequency analysis problems: close frequencies distinction, frequency estimation and Signal to Noise Ratio estimation. The results are presented, when possible, with real world audio signals.

Detecting single-trial EEG evoked potential using a wavelet domain linear mixed model: application to error potentials classification

Auteur(s): Torresani B.

(Article) Publié: Journal Of Neural Engineering, vol. 12 p.036013 (2015)

Ref HAL: hal-01161911_v1
Ref Arxiv: 1506.07627
DOI: 10.1088/1741-2560/12/3/036013
Ref. & Cit.: NASA ADS
Exporter : BibTex | endNote
Résumé:

Objective. The main goal of this work is to develop a model for multi-sensor signals such as MEG or EEG signals, that accounts for the inter-trial variability, suitable for corresponding binary classification problems. An important constraint is that the model be simple enough to handle small size and unbalanced datasets, as often encountered in BCI type experiments. Approach. The method involves linear mixed effects statistical model, wavelet transform and spatial filtering, and aims at the characterization of localized discriminant features in multi-sensor signals. After discrete wavelet transform and spatial filtering, a projection onto the relevant wavelet and spatial channels subspaces is used for dimension reduction. The projected signals are then decomposed as the sum of a signal of interest (i.e. discriminant) and background noise, using a very simple Gaussian linear mixed model. Main results. Thanks to the simplicity of the model, the corresponding parameter estimation problem is simplified. Robust estimates of class-covariance matrices are obtained from small sample sizes and an effective Bayes plug-in classifier is derived. The approach is applied to the detection of error potentials in multichannel EEG data, in a very unbalanced situation (detection of rare events). Classification results prove the relevance of the proposed approach in such a context. Significance. The combination of linear mixed model, wavelet transform and spatial filtering for EEG classification is, to the best of our knowledge, an original approach, which is proven to be effective. This paper improves on earlier results on similar problems, and the three main ingredients all play an important role.

Analyse discriminante matricielle descriptive. Application a l'étude de signaux EEG

Auteur(s): Torresani B.

Conference: Journées de statistique de la SFDS (Lille, FR, 2015-06-01)
Actes de conférence: , vol. p. ()

Ref HAL: hal-01161902_v1
Ref Arxiv: 1506.02927
Ref. & Cit.: NASA ADS
Exporter : BibTex | endNote
Résumé:

Nous nous intéressons à l'approche descriptive de l'analyse discriminante linéaire de données matricielles dans le cas binaire. Sous l'hypothèse de séparabilité de la variabilité des lignes de celle des colonnes, les combinaisons linéaires des lignes et des colonnes les plus discriminantes sont déterminées par la décomposition en valeurs singulières de la différence des moyennes des deux classes en munissant les espaces des lignes et des colonnes de la métrique de Mahalanobis. Cette approche permet d'obtenir des représentations des données dans des plans factoriels et de dégager des composantes discriminantes. Une application a des signaux d'électroencéphalographie multi-capteurs illustre la pertinence de la méthode.

Time-frequency and time-scale analysis of deformed stationary processes, with application to non-stationary sound modeling

Auteur(s): Omer H., Torresani B.

(Article) Publié: Applied And Computational Harmonic Analysis, vol. p. (2016)

Ref HAL: hal-01094835_v2
Ref Arxiv: 1510.08240
DOI: 10.1016/j.acha.2015.10.002
Ref. & Cit.: NASA ADS
Exporter : BibTex | endNote
Résumé:

A class of random non-stationary signals termed timbre×dynamics is introduced and studied. These signals are obtained by non-linear transformations of sta-tionary random gaussian signals, in such a way that the transformation can be approximated by translations in an appropriate representation domain. In such situations, approximate maximum likelihood estimation techniques can be de-rived, which yield simultaneous estimation of the transformation and the power spectrum of the underlying stationary signal. This paper focuses on the case of modulation and time warping of station-ary signals, and proposes and studies estimation algorithms (based on time-frequency and time-scale representations respectively) for these quantities of interest. The proposed approach is validated on numerical simulations on synthetic signals, and examples on real life car engine sounds.

A review of blind source separation in NMR spectroscopy

Auteur(s): Toumi Ichrak, Caldarelli Stefano, Torresani B.

(Article) Publié: Progress In Nuclear Magnetic Resonance Spectroscopy, vol. 81 p.37-64 (2014)

Ref HAL: hal-01060561_v1
DOI: 10.1016/j.pnmrs.2014.06.002
Exporter : BibTex | endNote
Résumé:

Fourier transform is the data processing naturally associated to most NMR experiments. Notable exceptions are Pulse Field Gradient and relaxation analysis, the structure of which is only partially suitable for FT. With the revamp of NMR of complex mixtures, fueled by analytical challenges such as metabolomics, alternative and more apt mathematical methods for data processing have been sought, with the aim of decomposing the NMR signal into simpler bits. Blind source separation is a very broad definition regrouping several classes of mathematical methods for complex signal decomposition that use no hypothesis on the form of the data. Developed outside NMR, these algorithms have been increasingly tested on spectra of mixtures. In this review, we shall provide an historical overview of the application of blind source separation methodologies to NMR, including methods specifically designed for the specificity of this spectroscopy.

Commentaires: 27 pages

Plus...

Menu

Rechercher