Effective data co-reduction for multimedia similarity search

Huang, Zi, Shen, Heng Tao, Liu, Jiajun and Zhou, Xiaofang (2011). Effective data co-reduction for multimedia similarity search. In: Proceedings of the 2011 ACM SIGMOD International Conference on Management of Data. 2011 ACM SIGMOD/PODS Conference, Athens, Greece, (1021-1032). 12-16 June 2011. doi:10.1145/1989323.1989430

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads

Author Huang, Zi
Shen, Heng Tao
Liu, Jiajun
Zhou, Xiaofang
Title of paper Effective data co-reduction for multimedia similarity search
Conference name 2011 ACM SIGMOD/PODS Conference
Conference location Athens, Greece
Conference dates 12-16 June 2011
Proceedings title Proceedings of the 2011 ACM SIGMOD International Conference on Management of Data   Check publisher's open access policy
Journal name Association for Computing Machinery. Special Interest Group on Management of Data. International Conference Proceedings   Check publisher's open access policy
Place of Publication New York, United States
Publisher ACM (Association for Computing Machinery) Press
Publication Year 2011
Sub-type Fully published paper
DOI 10.1145/1989323.1989430
Open Access Status
ISBN 9781450306614
ISSN 0730-8078
Start page 1021
End page 1032
Total pages 12
Collection year 2012
Language eng
Abstract/Summary Multimedia similarity search has been playing a critical role in many novel applications. Typically, multimedia objects are described by high-dimensional feature vectors (or points) which are organized in databases for retrieval. Although many high-dimensional indexing methods have been proposed to facilitate the search process, efficient retrieval over large, sparse and extremely high-dimensional databases remains challenging due to the continuous increases in data size and feature dimensionality. In this paper, we propose the first framework for Data Co-Reduction (DCR) on both data size and feature dimensionality. By utilizing recently developed co-clustering methods, DCR simultaneously reduces both size and dimensionality of the original data into a compact subspace, where lower bounds of the actual distances in the original space can be efficiently established to achieve fast and lossless similarity search in the filter-andrefine approach. Particularly, DCR considers the duality between size and dimensionality, and achieves the optimal coreduction which generates the least number of candidates for actual distance computations. We conduct an extensive experimental study on large and real-life multimedia datasets, with dimensionality ranging from 432 to 1936. Our results demonstrate that DCR outperforms existing methods significantly for lossless retrieval, especially in the presence of extremely high dimensionality.
Keyword Data co-reduction
High-dimensional indexing
Q-Index Code E1
Q-Index Status Confirmed Code
Institutional Status UQ

Version Filter Type
Citation counts: Scopus Citation Count Cited 14 times in Scopus Article | Citations
Google Scholar Search Google Scholar
Created: Mon, 15 Aug 2011, 17:23:25 EST by Dr Heng Tao Shen on behalf of School of Information Technol and Elec Engineering