BEGIN:VCALENDAR

VERSION:2.0

PRODID:-//wp-events-plugin.com//6.6.4.4//EN

TZID:Asia/Jerusalem

X-WR-TIMEZONE:Asia/Jerusalem
BEGIN:VEVENT

UID:0-464@aerospace.technion.ac.il

DTSTART;TZID=Asia/Jerusalem:20170109T163000

DTEND;TZID=Asia/Jerusalem:20170109T173000

DTSTAMP:20230603T191039Z

URL:https://aerospace.technion.ac.il/events/incorporating-data-association
 -within-belief-space-planning-for-robust-autonomous-navigation/

SUMMARY:Incorporating Data Association within Belief Space Planning for Rob
 ust Autonomous Navigation
DESCRIPTION:Lecturer:Antony Thomas\n Faculty:Department of Aerospace Engine
 ering\n Institute:Technion – Israel Institute of Technology\n Location:C
 lassroom 165\, ground floor\, Library\, Aerospace Eng.\n Zoom: \n Abstract
 : \n Details: \n Belief space planning (BSP) and perception are fundamenta
 l problems in robotics and artificial intelligence\, with applications inc
 luding autonomous navigation and active SLAM. State-of-the-art BSP approa
 ches assume that data association (DA)\, i.e. determining the correct corr
 espondence between the observations and the landmarks\, is given and perf
 ect. However\, real world environments are often ambiguous\, which in the 
 presence of different sources of uncertainty\, make perception a challeng
 ing task. For example\, an object might be similar in appearance from the 
 current viewpoint to another object\, while successfully matching images 
 from two different but similar in appearance places (e.g. buildings that l
 ook alike) would incorrectly indicate the two places as one.  An incorre
 ct DA can lead to catastrophic results\, e.g. a robot considering it is lo
 cated in a wrong aliased corridor. Consequently\, more advanced approache
 s\, known as robust perception\, are required. Yet\, existing robust perce
 ption approaches focus on the passive case where robot actions are extern
 ally determined\, while existing BSP methods assume data association to be
  given and perfect.\n\nIn this research we relax the above assumption and 
 incorporate reasoning regarding DA aspects within BSP\, while accounting f
 or different sources of uncertainty (imperfect sensing\, stochastic contr
 ol\, uncertain environment). We develop a data association aware belief sp
 ace planning (DA-BSP) approach that explicitly reasons about DA within be
 lief evolution while considering non-myopic planning and multi-modal belie
 fs represented by Gaussian Mixture Models (GMM). We envision such a frame
 work to provide robust active perception and active disambiguation capabi
 lities\, in particular while operating in ambiguous and perceptually alias
 ed environments. The approach is studied and proven effective using real-
 world experiments and synthetic simulations\, carried out at the Autonomou
 s Navigation and Perception Lab at the Technion.
CATEGORIES:Seminars
LOCATION:Classroom 165\, ground floor\, Library\, Aerospace Eng.

END:VEVENT

BEGIN:VTIMEZONE

TZID:Asia/Jerusalem

X-LIC-LOCATION:Asia/Jerusalem

BEGIN:STANDARD

DTSTART:20161030T010000

TZOFFSETFROM:+0300

TZOFFSETTO:+0200

TZNAME:IST

END:STANDARD

END:VTIMEZONE
END:VCALENDAR