BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:CET
BEGIN:DAYLIGHT
DTSTART:20210328T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20201025T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20201203T145508Z
UID:7B8646CA-A39C-4B61-817E-2807950E1A14
DTSTART;TZID=CET:20201203T141500
DTEND;TZID=CET:20201203T151500
DESCRIPTION:Face super-resolution (or face hallucination) refers to the tas
 k of recovering high-resolution facial images from corresponding low-resol
 ution inputs. Solutions to this task have important applications in face-o
 riented vision problems\, such as face editing\, face alignment\, 3D face 
 reconstruction\, face attribute estimation and most notably face recogniti
 on. Driven by advances in deep learning\, recent years have seen tremendou
 s progress in this area with contemporary deep face hallucination models a
 chieving formidable performance. In my presentation\, I will first talk ab
 out the recent progress in the field of face super-resolution\, review exi
 sting approaches and discuss the most important trends in the area. Next\,
  I will present our solution to the problem of face hallucination\, which 
 uses explicit identity constraints in addition to the common reconstructio
 n loss during model training and show how it compares to state-of-the-art 
 hallucination models from the literature. I will elaborate on the main lim
 itations of existing face super-resolution approaches and present challeng
 es that will need to be addressed in the future. Finally\, I will describe
  our approach to face recognition from low-resolution images that relies o
 n super-resolution and is shown to result in state-of-the-art performance 
 on a popular benchmark.\n\nCo-sponsored by: Roma Tre University\, Dept. of
  Engineering\, Section of Applied Electronics\n\nSpeaker(s): Vitomir Struc
 \, \n\nVirtual: https://events.vtools.ieee.org/m/248896
LOCATION:Virtual: https://events.vtools.ieee.org/m/248896
ORGANIZER:patrizio.campisi@uniroma3.it
SEQUENCE:5
SUMMARY:Face super-resolution in biometrics: Recent advances and future cha
 llenges
URL;VALUE=URI:https://events.vtools.ieee.org/m/248896
X-ALT-DESC:Description: <br /><p>Face super-resolution (or face hallucinati
 on) refers to the task of recovering high-resolution facial images from co
 rresponding low-resolution inputs. Solutions to this task have important a
 pplications in face-oriented vision problems\, such as face editing\, face
  alignment\, 3D face reconstruction\, face attribute estimation and most n
 otably face recognition. Driven by advances in deep learning\, recent year
 s have seen tremendous progress in this area with contemporary deep face h
 allucination models achieving formidable performance. In my presentation\,
  I will first talk about the recent progress in the field of face super-re
 solution\, review existing approaches and discuss the most important trend
 s in the area. Next\, I will present our solution to the problem of face h
 allucination\, which uses explicit identity constraints in addition to the
  common reconstruction loss during model training and show how it compares
  to state-of-the-art hallucination models from the literature. I will elab
 orate on the main limitations of existing face super-resolution approaches
  and present challenges that will need to be addressed in the future. Fina
 lly\, I will describe our approach to face recognition from low-resolution
  images that relies on super-resolution and is shown to result in state-of
 -the-art performance on a popular benchmark.</p>
END:VEVENT
END:VCALENDAR