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:1847@i2m.univ-amu.fr
DTSTART;TZID=Europe/Paris:20170630T140000
DTEND;TZID=Europe/Paris:20170630T150000
DTSTAMP:20170615T120000Z
URL:https://www.i2m.univ-amu.fr/evenements/low-rank-time-frequency-synthes
 is/
SUMMARY: (...): Low-rank time-frequency synthesis
DESCRIPTION:: Many single-channel signal decomposition techniques rely on a
  low-rank factorization of a time-frequency transform. In particular\, non
 negative matrix factorization (NMF) of the spectrogram – the (power) mag
 nitude of the short-time Fourier transform (STFT) – has been considered 
 in many audio applications. In this setting\, NMF with the Itakura-Saito d
 ivergence was shown to underly a generative Gaussian composite model (GCM)
  of the STFT\, a step forward from more empirical approaches based on ad-h
 oc transform and divergence specifications. Still\, the GCM is not yet a g
 enerative model of the raw signal itself\, but only of its STFT. The work 
 presented in this paper fills in this ultimate gap by proposing a novel si
 gnal synthesis model with low-rank time-frequency structure. In particular
 \, our new approach opens doors to multi-resolution representations\, that
  were not possible in the traditional NMF setting. We describe expectation
 -maximization algorithms for estimation in the new model and report audio 
 signal processing results with music decomposition and new approach to the
  compressive sampling inverse problem\, that exploits latent low-rank time
 -frequency structure instead of sparsity\, with superior results for the c
 onsidered data.http://webpages.lss.supelec.fr/perso/matthieu.kowalski/
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:DAYLIGHT
DTSTART:20170326T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR