BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:EE Distinguished Speakers Seminar: Geometric perspectives on deep 
 learning models
DTSTART:20200512T140000
DTEND:20200512T150000
DTSTAMP:20260604T004111Z
UID:f6afe8d63ec0c41f53f661034359e9749afe3f1e7db7165be5c1eb11
CATEGORIES:Conferences - Seminars
DESCRIPTION:Pascal Frossard has been a faculty at EPFL since 2003\, and he
  leads the Signal Processing Laboratory (LTS4) of the Electrical Engineer
 ing Institute. His research interests include network data analysis\, ima
 ge representation and understanding\, signal processing  and machine lear
 ning. Before joining EPFL\, he was a member of the research staff at the 
 IBM T. J. Watson Research Center\, Yorktown Heights\, NY\, USA.\nAbstract
 : Learning-based decision systems become state-of-the-art in numerous appl
 ication domains. In particular\, deep learning has developed as a standar
 d in vision and speech applications for example. Despite this success\, i
 mportant research questions remain open: the robustness of deep learning 
 systems is surprisingly low\, and the interpretation of their decisions is
  often challenging. We will overview the recent research developed at the
  LTS4 laboratory\, towards developing algorithms for measuring the robust
 ness of deep networks\, deriving geometric insights on their behaviour\, a
 nd designing more effective systems. \n\nZoom link: https://epfl.zoom.us
 /j/96097462787?pwd=MUE1R0NxODVQNUFSZ0lqdlA4RWF5UT09
LOCATION:Zoom
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR