@tensen

News Story Segmentation in Multiple Modalities

, , und . CBMI: 2009 INTERNATIONAL WORKSHOP ON CONTENT-BASED MULTIMEDIA INDEXING, Seite 25--32. 345 E 47TH ST, NEW YORK, NY 10017 USA, IEEE; European Assoc Signal Proc; Inst Commun Comp Syst; IEEE Ciruits & Syst Soc Syst; Weknowit, IEEE, (2009)International Workshop on Content-Based Multimedia Indexing, Chania, GREECE, JUN 03-05, 2009.

Zusammenfassung

In this paper, we describe an approach to segmenting news video based on the perceived shift in content using features spanning multiple modalities. We investigate a number of multimedia features, which serve as potential indicators of a change in story in order to determine which are the most effective. The efficacy of our approach is demonstrated by the performance of our prototype, where a number of feature combinations demonstrate an up to 18\% improvement in WindowDiff score above that of other state of the art story segmenters. In our investigation, there was no, one, clearly superior feature, rather the best segmentation results occurred when there was synergy between multiple features.

Beschreibung

ISI Web of Knowledge [v.4.8] - All Databases Full Record

Links und Ressourcen

Tags

Community

  • @tensen
  • @dblp
@tensens Tags hervorgehoben