A novel interactive image retrieval method based on LSH and SVM

Author

Tingting Dong ; Zhicheng Zhao

Author_Institution

Multimedia Commun. & Pattern Recognition Labs., Beijing Univ. of Posts & Telecommun., Beijing, China

fYear

2012

fDate

21-23 Sept. 2012

Firstpage

482

Lastpage

486

Abstract

Content-based image retrieval (CBIR) has attracted people´s attention for many years, while the semantic gap and curse of dimensionality are still two open questions of CBIR. In this paper, we propose a new interactive image retrieval method based on locality-sensitive hashing (LSH) and support vector machine (SVM): LSH is adopted to overcome the curse of dimensionality and a SVM-based relevance feedback (RF) scheme is introduced to shorten the semantic gap. The experimental results show the effectiveness of the proposed method.

Keywords

content-based retrieval; cryptography; image retrieval; relevance feedback; support vector machines; CBIR; LSH; SVM; content-based image retrieval; curse of dimensionality; interactive image retrieval; locality-sensitive hashing; relevance feedback; semantic gap; support vector machine; Feature extraction; Image retrieval; Indexes; Radio frequency; Support vector machines; Vectors; Image retrieval; LSH; Relevance feedback; SVM;

fLanguage

English

Publisher

ieee

Conference_Titel

Network Infrastructure and Digital Content (IC-NIDC), 2012 3rd IEEE International Conference on

Conference_Location

Beijing

Print_ISBN

978-1-4673-2201-0

Type

conf

DOI

10.1109/ICNIDC.2012.6418800

Filename

6418800