Title
Graph-based Selective Rank Fusion for Unsupervised Image Retrieval
Abstract
•A novel unsupervised method to combine the retrieval results from various features.•The method exploits a graph-based representation for a selective rank fusion.•Both correlation and effectiveness estimations are used to construct the graph.•Experimental evaluation involving various public datasets and several features.•High-effective results achieved in comparison with baselines and state-of-the-art.
Year
DOI
Venue
2020
10.1016/j.patrec.2020.03.032
Pattern Recognition Letters
Keywords
DocType
Volume
Content-based image retrieval,Unsupervised late fusion,Rank-aggregation,Correlation measure,Effectiveness estimation
Journal
135
ISSN
Citations 
PageRank 
0167-8655
0
0.34
References 
Authors
0
2
Name
Order
Citations
PageRank
Lucas Pascotti Valem175.80
Daniel Carlos Guimarães Pedronette230425.47