Image re-ranking using graph based spanning structures and reciprocal nearest neighbors

B. Mocanu, R. Tapu, T. Zaharia

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

Abstract

In this paper we propose a novel method to improve the performance of image retrieval at VLAD (Vector of Locally Aggregated Descriptors) level. Our re-ranking algorithm uses relational graphs and the top- & neighborhood candidates to adaptively modify images similarity scores. The method is effective and increases the accuracy, without relying on low-level information or features geometrical verification.

Original languageEnglish
Title of host publication2016 IEEE International Conference on Consumer Electronics, ICCE 2016
EditorsFrancisco J. Bellido, Nicholas C. H. Vun, Carsten Dolar, Daniel Diaz-Sanchez, Wing-Kuen Ling
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages437-438
Number of pages2
ISBN (Electronic)9781467383646
DOIs
Publication statusPublished - 10 Mar 2016
Externally publishedYes
Event2016 IEEE International Conference on Consumer Electronics, ICCE 2016 - Las Vegas, United States
Duration: 7 Jan 201611 Jan 2016

Publication series

Name2016 IEEE International Conference on Consumer Electronics, ICCE 2016

Conference

Conference2016 IEEE International Conference on Consumer Electronics, ICCE 2016
Country/TerritoryUnited States
CityLas Vegas
Period7/01/1611/01/16

Fingerprint

Dive into the research topics of 'Image re-ranking using graph based spanning structures and reciprocal nearest neighbors'. Together they form a unique fingerprint.

Cite this