Článek ve sborníku konference

CHMELAŘ Petr, BERAN Vítězslav, HEROUT Adam, HRADIŠ Michal, JURÁNEK Roman, LÁNÍK Aleš, MLÍCH Jozef, NAVRÁTIL Jan, ŘEZNÍČEK Ivo, ŽÁK Pavel a ZEMČÍK Pavel. Brno University of Technology at TRECVid 2008. In: Proceedings of TRECVID 2008. Gaithersburg: National Institute of Standards and Technology, 2008, s. 1-16.
Jazyk publikace:angličtina
Název publikace:Brno University of Technology at TRECVid 2008
Název (cs):Brno University of Technology at TRECVid 2008
Strany:1-16
Sborník:Proceedings of TRECVID 2008
Konference:2008 TRECVID Workshop
Místo vydání:Gaithersburg, US
Rok:2008
Vydavatel:National Institute of Standards and Technology
URL:http://www.fit.vutbr.cz/research/groups/graph/publi/2008/2008-Chmelar-TRECVID2008.pdf [PDF]
Klíčová slova
TRECVid 2008, Brno, Surveillance event detection pilot, Content-based copy detection pilot, High-level feature extraction, Video Search.
Anotace
In this paper we describe our experiments in all task of TRECVid 2008. This year, we have concentrated
mainly on the local (affine covariant) image features and its transformation into a search-able form for the
Content-based copy detection pilot together with the indexing and search techniques for the Search task and a
practical test of the background subtraction and trajectory generation algorithms for the Surveillance pilot.
In brief, we have submitted the following tasks:
    1. Surveillance event detection pilot. We have participated in the detection of the following events
         - PersonRuns, ObjectPut, ElevatorNoEntry and OpposingFlow. It has been based mainly on
         advanced masking and background subtractions and extracted trajectories.
    2. Content-based copy detection pilot. We have submitted one run based on search of the joint
         image features - global (color, texture) and local features (SIFT).
    3. High-level feature extraction. We have used two training methods based on SVM using color,
         texture and face image features. First only selected subset of the training data, second all the
         annotated data were used for the training.
    4. Search. We have performed two fully automatic IR experiments based on the text of the queries
         and ASR/MT provided by NIST and the data consumed by the High-level feature extraction task.
    5. Rushes summarization, to which is dedicated a separate paper [2].

BibTeX:
@INPROCEEDINGS{
   author = {Petr Chmelař and Vítězslav Beran and Adam Herout and Michal
	Hradiš and Roman Juránek and Aleš Láník and Jozef Mlích and
	Jan Navrátil and Ivo Řezníček and Pavel Žák and Pavel Zemčík},
   title = {Brno University of Technology at TRECVid 2008},
   pages = {1--16},
   booktitle = {Proceedings of TRECVID 2008},
   year = {2008},
   location = {Gaithersburg, US},
   publisher = {National Institute of Standards and Technology},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php.cs?id=8828}
}

Vaše IPv4 adresa: 54.167.207.16
Přepnout na IPv6 spojení

DNSSEC [dnssec]