BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//AMSE//Event Calendar//FR
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
UID:event-12017@amse-aixmarseille.fr
DTSTAMP:20260415T034143Z
CREATED:20260415T034143Z
LAST-MODIFIED:20260415T034143Z
STATUS:CONFIRMED
SEQUENCE:0
SUMMARY:phd seminar - Anita Salvador
DTSTART:20250325T104500Z
DTEND:20250325T111500Z
DESCRIPTION:Non-ischemic cardiomyopathies represent almost 40% of hospitali
 zed patients with heart failure. These pathologies exhibit high variability
  in clinical presentation and progression due to their multi-factorial natu
 re. Characterizing a patient’s health state requires gathering informatio
 n from different sources (MRI images\, patient record\, ECG\, etc.) and agg
 regating them to obtain a simpler representation. I am working on different
  models within the family of Multimodal Variational Autoencoders (VAEs) whi
 ch offer a probabilistic framework for learning joint latent representation
 s from heterogeneous data. In this presentation\, I will discuss two archit
 ectures for multimodal VAEs: alignment-based and fusion-based approaches. W
 hile the alignment model learns a latent space for each modality\, the fusi
 on model aggregates distributions from each modality into a unified represe
 ntation. However\, the choice of aggregation method raises theoretical chal
 lenges\, one of them being computing the Kullback-Leibler divergence betwee
 n latent distributions.\\n\\nContact: Philippine Escudié: philippine.escu
 die[at]univ-amu.frLucie Giorgi: lucie.giorgi[at]univ-amu.frKla Kouadio: kla
 .kouadio[at]univ-amu.frLola Soubeyrand: lola.soubeyrand[at]univ-amu.fr\n\n
 Plus d'informations: https://amse-aixmarseille.fr/en/events/anita-salvador-
 
LOCATION:MEGA - Salle Carine Nourry\, 424\, Chemin du Viaduc\, 13080 Aix-en
 -Provence
URL;VALUE=URI:https://amse-aixmarseille.fr/en/events/anita-salvador-0
CONTACT:Philippine Escudié:&nbsp\;philippine.escudie[at]univ-amu.frLucie G
 iorgi: lucie.giorgi[at]univ-amu.frKla Kouadio: kla.kouadio[at]univ-amu.frLo
 la Soubeyrand:&nbsp\;lola.soubeyrand[at]univ-amu.fr
TRANSP:OPAQUE
END:VEVENT
END:VCALENDAR