The uncertain representation ranking framework for concept-based video retrieval

Robin Aly, Aiden Doherty, Djoerd Hiemstra, Franciska M.G. de Jong, Alan F. Smeaton

  • 1 Citations

Abstract

Concept based video retrieval often relies on imperfect and uncertain concept detectors. We propose a general ranking framework to define effective and robust ranking functions, through explicitly addressing detector uncertainty. It can cope with multiple concept-based representations per video segment and it allows the re-use of effective text retrieval functions which are defined on similar representations. The final ranking status value is a weighted combination of two components: the expected score of the possible scores, which represents the risk-neutral choice, and the scores’ standard deviation, which represents the risk or opportunity that the score for the actual representation is higher. The framework consistently improves the search performance in the shot retrieval task and the segment retrieval task over several baselines in five TRECVid collections and two collections which use simulated detectors of varying performance.
Original languageUndefined
Pages (from-to)557-583
Number of pages27
JournalInformation retrieval
Volume16
Issue number5
DOIs
StatePublished - Oct 2013

Fingerprint

Detectors
ranking
score
representation
video
concept
function
risk
performance
uncertainty
text

Keywords

  • Representation Uncertainty · Concept-based Representation · Video Retrieval
  • IR-80911
  • METIS-287734
  • EWI-22070

Cite this

Aly, Robin; Doherty, Aiden; Hiemstra, Djoerd; de Jong, Franciska M.G.; Smeaton, Alan F. / The uncertain representation ranking framework for concept-based video retrieval.

In: Information retrieval, Vol. 16, No. 5, 10.2013, p. 557-583.

Research output: Scientific - peer-reviewArticle

@article{92b49fb2f06446c08d11a58256bf6f0c,
title = "The uncertain representation ranking framework for concept-based video retrieval",
abstract = "Concept based video retrieval often relies on imperfect and uncertain concept detectors. We propose a general ranking framework to define effective and robust ranking functions, through explicitly addressing detector uncertainty. It can cope with multiple concept-based representations per video segment and it allows the re-use of effective text retrieval functions which are defined on similar representations. The final ranking status value is a weighted combination of two components: the expected score of the possible scores, which represents the risk-neutral choice, and the scores’ standard deviation, which represents the risk or opportunity that the score for the actual representation is higher. The framework consistently improves the search performance in the shot retrieval task and the segment retrieval task over several baselines in five TRECVid collections and two collections which use simulated detectors of varying performance.",
keywords = "Representation Uncertainty · Concept-based Representation · Video Retrieval, IR-80911, METIS-287734, EWI-22070",
author = "Robin Aly and Aiden Doherty and Djoerd Hiemstra and {de Jong}, {Franciska M.G.} and Smeaton, {Alan F.}",
year = "2013",
month = "10",
doi = "10.1007/s10791-012-9207-y",
volume = "16",
pages = "557--583",
journal = "Information retrieval",
issn = "1386-4564",
publisher = "Springer Netherlands",
number = "5",

}

The uncertain representation ranking framework for concept-based video retrieval. / Aly, Robin; Doherty, Aiden; Hiemstra, Djoerd; de Jong, Franciska M.G.; Smeaton, Alan F.

In: Information retrieval, Vol. 16, No. 5, 10.2013, p. 557-583.

Research output: Scientific - peer-reviewArticle

TY - JOUR

T1 - The uncertain representation ranking framework for concept-based video retrieval

AU - Aly,Robin

AU - Doherty,Aiden

AU - Hiemstra,Djoerd

AU - de Jong,Franciska M.G.

AU - Smeaton,Alan F.

PY - 2013/10

Y1 - 2013/10

N2 - Concept based video retrieval often relies on imperfect and uncertain concept detectors. We propose a general ranking framework to define effective and robust ranking functions, through explicitly addressing detector uncertainty. It can cope with multiple concept-based representations per video segment and it allows the re-use of effective text retrieval functions which are defined on similar representations. The final ranking status value is a weighted combination of two components: the expected score of the possible scores, which represents the risk-neutral choice, and the scores’ standard deviation, which represents the risk or opportunity that the score for the actual representation is higher. The framework consistently improves the search performance in the shot retrieval task and the segment retrieval task over several baselines in five TRECVid collections and two collections which use simulated detectors of varying performance.

AB - Concept based video retrieval often relies on imperfect and uncertain concept detectors. We propose a general ranking framework to define effective and robust ranking functions, through explicitly addressing detector uncertainty. It can cope with multiple concept-based representations per video segment and it allows the re-use of effective text retrieval functions which are defined on similar representations. The final ranking status value is a weighted combination of two components: the expected score of the possible scores, which represents the risk-neutral choice, and the scores’ standard deviation, which represents the risk or opportunity that the score for the actual representation is higher. The framework consistently improves the search performance in the shot retrieval task and the segment retrieval task over several baselines in five TRECVid collections and two collections which use simulated detectors of varying performance.

KW - Representation Uncertainty · Concept-based Representation · Video Retrieval

KW - IR-80911

KW - METIS-287734

KW - EWI-22070

U2 - 10.1007/s10791-012-9207-y

DO - 10.1007/s10791-012-9207-y

M3 - Article

VL - 16

SP - 557

EP - 583

JO - Information retrieval

T2 - Information retrieval

JF - Information retrieval

SN - 1386-4564

IS - 5

ER -