Improving Similarity Search in Face-Images Data

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Similarity search involves finding all the face-images in a database,which are similar to a desired face-image, based on some distance measure.Comparing the desired face-image to all the face-images in a large dataset isprohibitively slow. If face-images can be placed in a metric space, search can besped up by using a metric data structure. In this work, we evaluate theperformance of range queries with metric data structures (LAESA, VPtree,DSAT, HDSAT1, HDSAT2, LC, RLC and GNAT) when the metric spaces areface-images data with the Euclidean distance. The experimental results showthat all data structures reduce the ratio between the number of distancescomputed and the database size. Moreover, the LAESA has the bestperformance in the majority of the experimental cases, but the RLC competeswith the other metric data structures, and has the best results when comparedwith the other dynamic metric data structures.
Original languageUnknown
Title of host publicationDELOS workshop proceedings series
PublisherDELOS digital library
Pages7-11
ISBN (Print)NONE
Publication statusPublished - 1 Jan 2009
EventSecond Workshop on Very Large Digital Libraries (VLDL 2009), 13th European Conference on Research and Advanced Technologies on Digital Libraries (ECDL 2009) -
Duration: 1 Jan 2009 → …

Conference

ConferenceSecond Workshop on Very Large Digital Libraries (VLDL 2009), 13th European Conference on Research and Advanced Technologies on Digital Libraries (ECDL 2009)
Period1/01/09 → …

Cite this