BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:Europe/Paris
X-WR-TIMEZONE:Europe/Paris
BEGIN:VEVENT
UID:5497@i2m.univ-amu.fr
DTSTART;TZID=Europe/Paris:20130328T140000
DTEND;TZID=Europe/Paris:20130328T150000
DTSTAMP:20241029T164523Z
URL:https://www.i2m.univ-amu.fr/evenements/m-kowalski-l2s-social-sparsity-
 application-to-audio-inpainting/
SUMMARY: (...): M. Kowalski (L2S): Social Sparsity: application to audio in
 painting.
DESCRIPTION:: Social Sparsity: application to audio inpainting.\n\nBy Matth
 ieu Kowalski\\\, L2S.\n\nAbstract:\nAudio inpainting problem is under cons
 ideration\\\, using iterative\nthresholding algorithms build on the "socia
 l sparsity" principle.\nFirst\\\, we present new shrinkage operators\\\, a
 llowing one to take into\naccount the neighborhood of time-frequency coeff
 icients. Then\\\, the\naudio declipping problem is formulated as a unconst
 rained convex\noptimization problem\\\, but taking into account an inporta
 nt hypothesis\nof audio declipping: reconstructed samples must be greater 
 than the\nclipping threshold. The structured thresholding operators\\\, su
 ch as the\nwindowed group-Lasso or the persistent empirical wiener\\\, are
  embedded\ninto iterative algorithms\\\, and we show on experimental resul
 ts the SNR\nimprovement compared to a more conventional Lasso. We also com
 pare the\nresults to the state of the art audio declipping.\n\nDownload sl
 ides
CATEGORIES:Séminaire,Signal et Apprentissage
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:STANDARD
DTSTART:20121028T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
END:STANDARD
END:VTIMEZONE
END:VCALENDAR