English

Using Apache Lucene to Search Vector of Locally Aggregated Descriptors

Computer Vision and Pattern Recognition 2016-04-20 v1 Information Retrieval

Abstract

Surrogate Text Representation (STR) is a profitable solution to efficient similarity search on metric space using conventional text search engines, such as Apache Lucene. This technique is based on comparing the permutations of some reference objects in place of the original metric distance. However, the Achilles heel of STR approach is the need to reorder the result set of the search according to the metric distance. This forces to use a support database to store the original objects, which requires efficient random I/O on a fast secondary memory (such as flash-based storages). In this paper, we propose to extend the Surrogate Text Representation to specifically address a class of visual metric objects known as Vector of Locally Aggregated Descriptors (VLAD). This approach is based on representing the individual sub-vectors forming the VLAD vector with the STR, providing a finer representation of the vector and enabling us to get rid of the reordering phase. The experiments on a publicly available dataset show that the extended STR outperforms the baseline STR achieving satisfactory performance near to the one obtained with the original VLAD vectors.

Keywords

Cite

@article{arxiv.1604.05576,
  title  = {Using Apache Lucene to Search Vector of Locally Aggregated Descriptors},
  author = {Giuseppe Amato and Paolo Bolettieri and Fabrizio Falchi and Claudio Gennaro and Lucia Vadicamo},
  journal= {arXiv preprint arXiv:1604.05576},
  year   = {2016}
}

Comments

In Proceedings of the 11th Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2016) - Volume 4: VISAPP, p. 383-392