Cristin-resultat-ID: 229887
Sist endret: 21. januar 2015 15:27
Resultat
Vitenskapelig foredrag
2007

Capturing Image Semantics through Collection Information

Bidragsytere:
  • Randi Karlsen og
  • Joan C. Nordbotten

Presentasjon

Navn på arrangementet: VERDIKT
Sted: Trondheim
Dato fra: 29. oktober 2007
Dato til: 30. oktober 2007

Arrangør:

Arrangørnavn: NFR

Om resultatet

Vitenskapelig foredrag
Publiseringsår: 2007

Klassifisering

Vitenskapsdisipliner

Informasjons- og kommunikasjonsteknologi

Beskrivelse Beskrivelse

Tittel

Capturing Image Semantics through Collection Information

Sammendrag

A rapidly increasing number of image collections are available over the Internet for browsing and searching by information seekers. Unfortunately, current techniques for image retrieval have shortcomings that make it difficult to search for images based on a semantic understanding of what the image is about. We believe that an extended use of image context information can enhance automatic identification of image semantics and thus may contribute to closing the semantic gap between user needs and the current capabilities of image retrieval systems. In this paper we focus on using Usage Context information to enhance the amount of image metadata. A Usage Context is the environment in which the image is used, typically in an image collection or as an illustration in a text document. In particular, we focus on how multiple situational contexts can be combined to provide enhanced knowledge of the semantics of individual images. We propose a language, CTXT, for context specification and demonstrate how Usage Context information can be represented so that the characteristics of each context type can be sustained and made available for use in image retrieval.

Bidragsytere

Randi Karlsen

  • Tilknyttet:
    Forfatter
    ved Inst. for informasjon og medievitenskap ved Universitetet i Bergen

Joan C. Nordbotten

  • Tilknyttet:
    Forfatter
    ved Inst. for informasjon og medievitenskap ved Universitetet i Bergen
1 - 2 av 2