Image2Text (I2T) tagged video technology

9 views
Skip to first unread message

Glenn

unread,
Jun 2, 2010, 10:26:46 PM6/2/10
to The ALISA Community
This article discusses work done by the Univ of CA and ObjectVideo
(Reston, VA) on an Image2Text (I2T) technology. Part of the project
included the tagging of a large data set at the Lotus Hill Inst
(China). Objects in various video scenes were tagged by humans and
then used to train the I2T system.

Article: http://www.technologyreview.com/computing/25439/?a=f

Some of their tagged data is available for free (with registration)
at:
http://www.imageparsing.com/FreeDataOutline.html

Glenn

unread,
Jul 13, 2010, 8:12:32 PM7/13/10
to The ALISA Community
DARPA Video and Image Retrieval and Analysis Tool

Mission : The overall goal of the Video and Image Retrieval and
Analysis Tool (VIRAT) program is to produce a scalable and extensible
end-to-end system that enables military analysts to obtain greater
value from aerial video collected in combat environments.

http://www.darpa.mil/ipto/programs/virat/virat.asp

Glenn

unread,
Jul 13, 2010, 8:20:52 PM7/13/10
to The ALISA Community
Algorithms for behavior recognition in wide-area video of urban
environments by Georgiy Levchuk

Automated analysis and identification of suspect behaviors and
functions of buildings, people, and vehicles are needed for real-time,
actionable intelligence.

http://spie.org/x40851.xml?highlight=x2412&ArticleID=x40851

Glenn

unread,
Jul 19, 2010, 7:33:46 PM7/19/10
to The ALISA Community
Forwarded from Bob Fisher
Person tracking data set available
--------------------------------------------------------------------------------
ANN: large overhead camera person tracking dataset

We have been contructing a video dataset of detected people walking
through
the atrium of a building, where the data comes from an overhead
camera.
Because of the data volume, we are not recording the video images, but
are doing target detection on the images and save the detection
records.
The records include a bounding box and a colour histogram for each
detected
target. Because of lighting, camera jitter and stationary objects,
there are
detections of non-people, too. After detection, we also have applied
our person tracker, recording the trajectories as both lists of
detections
and splines.

By July 19, 2010, there were 29 million target detections, of which
an estimated 8.3 million were real targets, resulting in 96K observed
trajectories. And we're still acquiring data!

You might find it interesting to use the data. It's at:
http://homepages.inf.ed.ac.uk/rbf/FORUMTRACKING/

The data was acquired in the Informatics Forum, the main building of
the
School of Informatics at the University of Edinburgh.

By the way, there are over 120 public image and video datasets
on CVonline at:
http://homepages.inf.ed.ac.uk/cgi/rbf/CVONLINE/entries.pl?TAG363

Best wishes, Bob

================

Prof. Robert B. Fisher, School of Informatics, Univ. of Edinburgh
1.26 Informatics Forum, 10 Crichton St, Edinburgh EH8 9AB, UK
+44-(131)-651-3441 (direct line), +44-(131)-651-3443 (secretary)
Fax: +44-(131)-650-6899 E-MAIL: r.b.fisher A@T ed.ac.uk
http://www.inf.ed.ac.uk/people/staff/Robert_Fisher.html

The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
Reply all
Reply to author
Forward
0 new messages