Generative Probabilistic Models

T.H.W. Westerveld, Franciska M.G. de Jong

Research output: Chapter in Book/Report/Conference proceedingChapterAcademicpeer-review

Abstract

Many content-based multimedia retrieval tasks can be seen as decision theory problems. Clearly, this is the case for classification tasks, like face detection, face recognition, or indoor/outdoor classification. In all these cases a system has to decide whether an image (or video) belongs to one class or another (respectively face or no face; face A, B, or C; and indoor or outdoor). Even the ad hoc retrieval tasks, where the goal is to find relevant documents given a description of an information need, can be seen as a decision theory problem: documents can be classified into relevant and non-relevant classes, or we can treat each of the documents in the collection as a separate class, and classify a query as belonging to one of these. In all these settings, a probabilistic approach seems natural: an image is assigned to the class with the highest probability.3 If some misclassifications are more severe than others, a decision theoretic approach should be taken, and images should be assigned to the class with lowest risk.
Original languageUndefined
Title of host publicationMultimedia Retrieval
EditorsHenk Blanken, H.E. Blok, A.P. de Vries, L. Feng
Place of PublicationBerlin Heidelberg
PublisherSpringer
Pages177-198
Number of pages22
ISBN (Print)978-3-540-72894-8
DOIs
Publication statusPublished - 2007

Publication series

NameData-Centric Systems and Applications
PublisherSpringer Verlag
Number2

Keywords

  • METIS-241569
  • EWI-9624
  • IR-63988

Cite this

Westerveld, T. H. W., & de Jong, F. M. G. (2007). Generative Probabilistic Models. In H. Blanken, H. E. Blok, A. P. de Vries, & L. Feng (Eds.), Multimedia Retrieval (pp. 177-198). [10.1007/978-3-540-72895-5_6] (Data-Centric Systems and Applications; No. 2). Berlin Heidelberg: Springer. https://doi.org/10.1007/978-3-540-72895-5_6
Westerveld, T.H.W. ; de Jong, Franciska M.G. / Generative Probabilistic Models. Multimedia Retrieval. editor / Henk Blanken ; H.E. Blok ; A.P. de Vries ; L. Feng. Berlin Heidelberg : Springer, 2007. pp. 177-198 (Data-Centric Systems and Applications; 2).
@inbook{10fc4b4383ef454a873c82773a4f1f13,
title = "Generative Probabilistic Models",
abstract = "Many content-based multimedia retrieval tasks can be seen as decision theory problems. Clearly, this is the case for classification tasks, like face detection, face recognition, or indoor/outdoor classification. In all these cases a system has to decide whether an image (or video) belongs to one class or another (respectively face or no face; face A, B, or C; and indoor or outdoor). Even the ad hoc retrieval tasks, where the goal is to find relevant documents given a description of an information need, can be seen as a decision theory problem: documents can be classified into relevant and non-relevant classes, or we can treat each of the documents in the collection as a separate class, and classify a query as belonging to one of these. In all these settings, a probabilistic approach seems natural: an image is assigned to the class with the highest probability.3 If some misclassifications are more severe than others, a decision theoretic approach should be taken, and images should be assigned to the class with lowest risk.",
keywords = "METIS-241569, EWI-9624, IR-63988",
author = "T.H.W. Westerveld and {de Jong}, {Franciska M.G.}",
note = "10.1007/978-3-540-72895-5_6",
year = "2007",
doi = "10.1007/978-3-540-72895-5_6",
language = "Undefined",
isbn = "978-3-540-72894-8",
series = "Data-Centric Systems and Applications",
publisher = "Springer",
number = "2",
pages = "177--198",
editor = "Henk Blanken and H.E. Blok and {de Vries}, A.P. and L. Feng",
booktitle = "Multimedia Retrieval",

}

Westerveld, THW & de Jong, FMG 2007, Generative Probabilistic Models. in H Blanken, HE Blok, AP de Vries & L Feng (eds), Multimedia Retrieval., 10.1007/978-3-540-72895-5_6, Data-Centric Systems and Applications, no. 2, Springer, Berlin Heidelberg, pp. 177-198. https://doi.org/10.1007/978-3-540-72895-5_6

Generative Probabilistic Models. / Westerveld, T.H.W.; de Jong, Franciska M.G.

Multimedia Retrieval. ed. / Henk Blanken; H.E. Blok; A.P. de Vries; L. Feng. Berlin Heidelberg : Springer, 2007. p. 177-198 10.1007/978-3-540-72895-5_6 (Data-Centric Systems and Applications; No. 2).

Research output: Chapter in Book/Report/Conference proceedingChapterAcademicpeer-review

TY - CHAP

T1 - Generative Probabilistic Models

AU - Westerveld, T.H.W.

AU - de Jong, Franciska M.G.

N1 - 10.1007/978-3-540-72895-5_6

PY - 2007

Y1 - 2007

N2 - Many content-based multimedia retrieval tasks can be seen as decision theory problems. Clearly, this is the case for classification tasks, like face detection, face recognition, or indoor/outdoor classification. In all these cases a system has to decide whether an image (or video) belongs to one class or another (respectively face or no face; face A, B, or C; and indoor or outdoor). Even the ad hoc retrieval tasks, where the goal is to find relevant documents given a description of an information need, can be seen as a decision theory problem: documents can be classified into relevant and non-relevant classes, or we can treat each of the documents in the collection as a separate class, and classify a query as belonging to one of these. In all these settings, a probabilistic approach seems natural: an image is assigned to the class with the highest probability.3 If some misclassifications are more severe than others, a decision theoretic approach should be taken, and images should be assigned to the class with lowest risk.

AB - Many content-based multimedia retrieval tasks can be seen as decision theory problems. Clearly, this is the case for classification tasks, like face detection, face recognition, or indoor/outdoor classification. In all these cases a system has to decide whether an image (or video) belongs to one class or another (respectively face or no face; face A, B, or C; and indoor or outdoor). Even the ad hoc retrieval tasks, where the goal is to find relevant documents given a description of an information need, can be seen as a decision theory problem: documents can be classified into relevant and non-relevant classes, or we can treat each of the documents in the collection as a separate class, and classify a query as belonging to one of these. In all these settings, a probabilistic approach seems natural: an image is assigned to the class with the highest probability.3 If some misclassifications are more severe than others, a decision theoretic approach should be taken, and images should be assigned to the class with lowest risk.

KW - METIS-241569

KW - EWI-9624

KW - IR-63988

U2 - 10.1007/978-3-540-72895-5_6

DO - 10.1007/978-3-540-72895-5_6

M3 - Chapter

SN - 978-3-540-72894-8

T3 - Data-Centric Systems and Applications

SP - 177

EP - 198

BT - Multimedia Retrieval

A2 - Blanken, Henk

A2 - Blok, H.E.

A2 - de Vries, A.P.

A2 - Feng, L.

PB - Springer

CY - Berlin Heidelberg

ER -

Westerveld THW, de Jong FMG. Generative Probabilistic Models. In Blanken H, Blok HE, de Vries AP, Feng L, editors, Multimedia Retrieval. Berlin Heidelberg: Springer. 2007. p. 177-198. 10.1007/978-3-540-72895-5_6. (Data-Centric Systems and Applications; 2). https://doi.org/10.1007/978-3-540-72895-5_6