dc.contributor.author | Avrithis, Yannis S. | en |
dc.contributor.author | Emiris, Ioannis Z. | en |
dc.contributor.author | Samaras, George S. | en |
dc.creator | Avrithis, Yannis S. | en |
dc.creator | Emiris, Ioannis Z. | en |
dc.creator | Samaras, George S. | en |
dc.date.accessioned | 2019-11-13T10:38:22Z | |
dc.date.available | 2019-11-13T10:38:22Z | |
dc.date.issued | 2016 | |
dc.identifier.isbn | 978-1-4503-4123-3 | |
dc.identifier.uri | http://gnosis.library.ucy.ac.cy/handle/7/53588 | |
dc.description.abstract | We propose a new data-structure, the generalized random-ized k-d forest, or k-d GeRaF, for approximate nearest neigh-bor searching in high dimensions. In particular, we intro-duce new randomization techniques to specify a set of in-dependently constructed trees where search is performed simultaneously, hence increasing accuracy. We omit back-tracking, and we optimize distance computations. We re-lease public domain software GeRaF and we compare it to existing implementations of state-of-the-art methods. Ex-perimental results on SIFT and GIST visual descriptors, in-dicate that our method is the method of choice in dimen-sions around 1,000, and probably up to 10,000, and datasets of cardinality up to a few hundred thousands or even one million. For instance, we handle a real dataset of 106 GIST images represented in 960 dimensions with a query time of leb than 1 sec on average and 90% responses being true nearest neighbors. © 2016 ACM. | en |
dc.publisher | Association for Computing Machinery | en |
dc.source | ACM International Conference Proceeding Series | en |
dc.source | 33rd Computer Graphics International Conference, CGI 2016 | en |
dc.source.uri | https://www.scopus.com/inward/record.uri?eid=2-s2.0-84978879979&doi=10.1145%2f2949035.2949042&partnerID=40&md5=c7acc1b78b00331fccc0c2ec9ee68cd4 | |
dc.subject | Forestry | en |
dc.subject | Computer graphics | en |
dc.subject | Open source software | en |
dc.subject | Color image processing | en |
dc.subject | GIST images | en |
dc.subject | High Dimen-Sion | en |
dc.subject | Image Search | en |
dc.subject | Open Software | en |
dc.subject | Randomized Tree | en |
dc.subject | Randomized trees | en |
dc.subject | Space Partition | en |
dc.title | High-dimensional visual similarity search: K-d generalized randomized forests [extended abstract] | en |
dc.type | info:eu-repo/semantics/conferenceObject | |
dc.identifier.doi | 10.1145/2949035.2949042 | |
dc.description.volume | 28-June-01-July-2016 | en |
dc.description.startingpage | 25 | |
dc.description.endingpage | 28 | |
dc.author.faculty | 002 Σχολή Θετικών και Εφαρμοσμένων Επιστημών / Faculty of Pure and Applied Sciences | |
dc.author.department | Τμήμα Πληροφορικής / Department of Computer Science | |
dc.type.uhtype | Conference Object | en |
dc.description.notes | <p>Sponsors: Foundation for Research and Technology - Hellas (FORTH) | en |
dc.description.notes | Conference code: 122396</p> | en |