Computing at Glasgow University
Paper ID: 8960

An Efficient Indexing Struture for Multimedia Data
Urruty,T. Djeraba,C. Jose,J.M.

Publication Type: Conference Proceedings
Appeared in: ACM MIR 08
Page Numbers :
Publisher: N/A
Year: 2008

In the last few years, the increase of online video has challenged research in the field of video information retrieval. Video search engines have become common on the Internet and require the use of powerful tools for fast access to data. However the representation of multimedia data as video shot or keyframe with visual features requires the use of a multidimensional space and indexing structures face the well known ``curse of dimensionality''. In this paper, we propose a new indexing structure that combines a clustering algorithm using random projections and a recursive multidimensional indexing structure. In our experiments, we study the effeciency and the effectiveness of our indexing structure using visual features of video shots of TRECVID database. We compare our proposed structure with other state-of-the-art methods.

Keywords: Clustering, Indexing Structure, Multimedia

Bibtex entry Endnote XML