Universität Karlsruhe (TH) at TRECVID 2007

H. K. Ekenel, M. Fischer, H. Gao, K. Kilgour, J. S. Marcos, R. Stiefelhagen

Research output: Contribution to conferencePaperpeer-review

9 Citations (Scopus)

Abstract

In this paper, we present the systems developed by the interACT Research Center at Universiẗat Karlsruhe (TH) for the TRECVID 2007 evaluation. Our first participation in the TRECVID evaluation includes the shot boundary detection and high-level feature detection tasks. The shot boundary detection system contains four separate detectors, one for each type of shot boundary we try to detect: cuts, fast dissolves, fade in / fade outs and dissolves. High-level feature detection utilizes both visual and textual cues. It uses color, texture and edge-based features for visual processing and relative term frequency-based features for text processing. All the systems we have developed have been trained on the common development data.

Original languageEnglish
Publication statusPublished - 2007
Externally publishedYes
EventTREC Video Retrieval Evaluation, TRECVID 2007 - Gaithersburg, MD, United States
Duration: 5 Nov 20076 Nov 2007

Conference

ConferenceTREC Video Retrieval Evaluation, TRECVID 2007
Country/TerritoryUnited States
CityGaithersburg, MD
Period5/11/076/11/07

Fingerprint

Dive into the research topics of 'Universität Karlsruhe (TH) at TRECVID 2007'. Together they form a unique fingerprint.

Cite this