Majorization-Minimization for Sparse Nonnegative Matrix Factorization With the $\beta$-Divergence

Autor: Arthur Marmin, José Henrique de Morais Goulart, Cédric Févotte

Publikováno v: IEEE Transactions on Signal Processing
IEEE Transactions on Signal Processing, 2023, 71, pp.1435-1447. ⟨10.1109/TSP.2023.3266939⟩

This article introduces new multiplicative updates for nonnegative matrix factorization with the $\beta$-divergence and sparse regularization of one of the two factors (say, the activation matrix). It is well known that the norm of the other factor (

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::c1a9747717bc488f2de2a3769b5f23a3
https://doi.org/10.1109/tsp.2023.3266939

Wavelets in the Deep Learning Era

Autor: Zaccharie Ramzi, Kevin Michalewicz, Jean-Luc Starck, Thomas Moreau, Philippe Ciuciu

Publikováno v: Journal of Mathematical Imaging and Vision
Journal of Mathematical Imaging and Vision, 2022, 65, ⟨10.1007/s10851-022-01123-w⟩

International audience; Sparsity-based methods, such as wavelets, have been the state of the art for more than 20 years for inverse problems before being overtaken by neural networks. In particular, U-nets have proven to be extremely effective. Their

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::db75ec47d468173598853300e3f709fe
https://doi.org/10.1007/s10851-022-01123-w

Stability of Unfolded Forward-Backward to Perturbations in Observed Data

Autor: de Valle, Cécile, Centofanti, Ezequiel, Chouzenoux, Emilie, Pesquet, Jean-Christophe

Publikováno v: Proceedings of the 31st European Signal Processing Conference (EUSIPCO 2023)
EUSIPCO 2023-31st European Signal Processing Conference
EUSIPCO 2023-31st European Signal Processing Conference, Sep 2023, Helsinki, Finland

International audience; We consider a neural network architecture to solve inverse problems, which is built by unfolding a forwardbackward algorithm. This algorithm is based on the minimization of an objective function which corresponds to a penalize

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=od______3379::e100632d89f068b611754d5aa1e43769
https://inria.hal.science/hal-04165350

Compression quasi sans perte d'images satellites par filtrage de résidus

Autor: Bacchus, Pascal, Roumy, Aline, Fraisse, Renaud, Guillemot, Christine

Publikováno v: Actes du Gretsi 2023
GRETSI 2023-XXIXème Colloque Francophone de Traitement du Signal et des Images
GRETSI 2023-XXIXème Colloque Francophone de Traitement du Signal et des Images, Aug 2023, Grenoble, France. pp.1-5
XXIXème Colloque Francophone de Traitement du Signal et des Images (GRETSI'23)
XXIXème Colloque Francophone de Traitement du Signal et des Images (GRETSI'23), Aug 2023, Grenoble, France

National audience; We describe an end-to-end trainable neural network for satellite image compression. We first introduce a cost functioncombining a perceptual loss and the classical mean square error that significantly improves the performance of th

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::7b25b5e340e5956e9f4b5eeaefa982ec
https://hal.science/hal-04156801

Learning Multilingual Expressive Speech Representation for Prosody Prediction without Parallel Data

Autor: Duret, Jarod, Parcollet, Titouan, Estève, Yannick

Publikováno v: Speech Synthesis Workshop (SSW)
Speech Synthesis Workshop (SSW), Aug 2023, Grenoble, France

International audience; We propose a method for speech-to-speech emotionpreserving translation that operates at the level of discrete speech units. Our approach relies on the use of multilingual emotion embedding that can capture affective informatio

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::55f52d95e02ae2ae45f98bbab0275ff2
https://hal.science/hal-04144850/file/main.pdf