BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20180325T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20171029T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260425T092610Z
UID:5a20074410a91308565533@ist.ac.at
DTSTART:20171206T113000
DTEND:20171206T123000
DESCRIPTION:Speaker: Torsten Sattler\nhosted by Christoph Lampert\nAbstract
 : 3D scene perception is a key ability for robots\, as well as for any typ
 e of intelligent system designed to operate in the real world. Among 3D sc
 ene perception algorithms\, methods for 3D mapping reconstruct 3D models o
 f scenes from camera images. Visual localization techniques in turn use th
 ese maps to determine the position and orientation of one or more cameras 
 in the world. Visual localization and mapping are thus fundamental problem
 s that need to be solved reliably and robustly in order to enable autonomo
 us agents such as self-driving cars or drones.At the same time\, localizat
 ion and mapping algorithms are key technologies for Mixed and Augmented Re
 ality applications.Over the last years and decades\, tremendous progress h
 as been made in the area of 3D Computer Vision\, including impressive resu
 lts for localization and mapping. Still\, localization and mapping techniq
 ues can be rather brittle in challenging scenarios that are highly relevan
 t for practical applications. This talk gives an overview over these chall
 enges and explains how a higher-level understanding of the environment can
  help to solve some of them. In particular\, I will present algorithms for
  localization and 3D reconstruction that rely on semantic information. Thi
 s higher level of abstraction allows them to succeed under challenging con
 ditions that could not be handled by previous work relying on purely photo
 metric or geometric cues. I will then outline how these techniques can be 
 extended to tackle a certain family of open problems. I will finally concl
 ude the talk with a set of examples showing that algorithms for 3D scene p
 erception will need to become even smarter in order to allow complex scene
  interactions for robots and other types of intelligent systems.
LOCATION:Mondi Seminar Room 3\, Central Building\, ISTA
ORGANIZER:abonvent@ist.ac.at
SUMMARY:Torsten Sattler: The Semantics of Visual Localization and Mapping
URL:https://talks-calendar.ista.ac.at/events/984
END:VEVENT
END:VCALENDAR