P A G E 3 I S S U E 1
Online demo for lecture videos
Participation in the TRECVID benchmarking activity
Communication and Dissemination activities
ments this year were performed on a
set of Internet Archive videos totaling
about 600 hours of video duraon, and
using 30 dierent queries. Our fully au-
tomac runs performed very well in this
challenging task, compared to the runs
of the other parcipang instuons
from all over the world. Specically, our
best run was ranked 2nd-best, achieving
an inferred average precision of 0.051
(compared to 0.054 reached by the best
-performing parcipant in the fully-
automac category, and 0.040 reached
by the 3rd best-performing one). Inter-
esngly, our fully automac runs also
compared favorably to the manually-
assisted runs that were submied to
AVS: with an inferred average precision
of 0.051, our best fully automac run
also outperformed the runs of all but
one parcipant in the manually-assisted
run category. We also had very good
results in the event-based annotaon
task (MED), where we tested our ma-
chine learning techniques for video an-
notaon, using a variable number of
training samples. Our parcipaon in
the AVS and MED tasks this year was
jointly supported by MOVING and by
another H2020 EU project, InVID.
MOVING, via its
consorum mem-
ber CERTH, suc-
cessfully parci-
pated in the Ad-
hoc Video Search
(AVS) and event-
based annotaon
(MED) tasks of
TRECVID 2016. The AVS task aempts to
model the end-user video search use-
case, where the user is looking for seg-
ments of video containing persons, ob-
jects, acvies, locaons, etc. and com-
binaons of the former. The experi-
Interview about MOVING to ZBW-mediatalk
ZBW’s interview about MOVING
The interacve
online demo for
lecture videos
such as automacally detected shots,
scenes, and visual concepts. You can
access the demo at: hp://mulmedia2
.i.gr/moving-project/lecture-video-link
ing-demo/results.html (best viewed wi-
th Firefox).
MOVING partner CERTH released an
interacve online demo linking lecture
videos, using general purpose concepts
that were produced from textual analy-
sis of their transcripts, with non-lecture
videos, using their visual analysis results
low the link:
hps://www.zbw-mediatalk.eu/2016/0
7/science-2-0-research-project-moving-
big-data-analyses-for-non-computer-sci
ensts/
Prof. Dr. Ansgar Schrep illustrated the
project details for ZBW-mediatalk, a
blog about technologies, services and
innovaons for libraries and media
companies. For more details please fol-