Integrating multiple global and local features by product sparse coding for image retrieval

Li Tian, Qi Jia, Seiichiro Kamata

Research output: Contribution to journalArticle

Abstract

In this study, we propose a simple, yet general and powerful framework of integrating multiple global and local features by Product Sparse Coding (PSC) for image retrieval. In our framework, multiple global and local features are extracted from images and then are transformed to Trimmed-Root (TR)-features. After that, the features are encoded into compact codes by PSC. Finally, a two-stage ranking strategy is proposed for indexing in retrieval. We make three major contributions in this study. First, we propose TR representation of multiple image features and show that the TR representation offers better performance than the original features. Second, the integrated features by PSC is very compact and effective with lower complexity than by the standard sparse coding. Finally, the two-stage ranking strategy can balance the efficiency and memory usage in storage. Experiments demonstrate that our compact image representation is superior to the state-of-the-art alternatives for large-scale image retrieval.

Original languageEnglish
Pages (from-to)731-738
Number of pages8
JournalIEICE Transactions on Information and Systems
VolumeE99D
Issue number3
DOIs
Publication statusPublished - 2016 Mar 1

Fingerprint

Image retrieval
Byproducts
Data storage equipment
Experiments

Keywords

  • Image representation
  • Image retrieval
  • Product Sparse Coding (PSC)
  • Ranking strategy
  • Trimmed-Root (TR)-feature

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Software
  • Artificial Intelligence
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition

Cite this

Integrating multiple global and local features by product sparse coding for image retrieval. / Tian, Li; Jia, Qi; Kamata, Seiichiro.

In: IEICE Transactions on Information and Systems, Vol. E99D, No. 3, 01.03.2016, p. 731-738.

Research output: Contribution to journalArticle

@article{47a49f6690864da9a314dd3d5eb99883,
title = "Integrating multiple global and local features by product sparse coding for image retrieval",
abstract = "In this study, we propose a simple, yet general and powerful framework of integrating multiple global and local features by Product Sparse Coding (PSC) for image retrieval. In our framework, multiple global and local features are extracted from images and then are transformed to Trimmed-Root (TR)-features. After that, the features are encoded into compact codes by PSC. Finally, a two-stage ranking strategy is proposed for indexing in retrieval. We make three major contributions in this study. First, we propose TR representation of multiple image features and show that the TR representation offers better performance than the original features. Second, the integrated features by PSC is very compact and effective with lower complexity than by the standard sparse coding. Finally, the two-stage ranking strategy can balance the efficiency and memory usage in storage. Experiments demonstrate that our compact image representation is superior to the state-of-the-art alternatives for large-scale image retrieval.",
keywords = "Image representation, Image retrieval, Product Sparse Coding (PSC), Ranking strategy, Trimmed-Root (TR)-feature",
author = "Li Tian and Qi Jia and Seiichiro Kamata",
year = "2016",
month = "3",
day = "1",
doi = "10.1587/transinf.2015EDP7337",
language = "English",
volume = "E99D",
pages = "731--738",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "3",

}

TY - JOUR

T1 - Integrating multiple global and local features by product sparse coding for image retrieval

AU - Tian, Li

AU - Jia, Qi

AU - Kamata, Seiichiro

PY - 2016/3/1

Y1 - 2016/3/1

N2 - In this study, we propose a simple, yet general and powerful framework of integrating multiple global and local features by Product Sparse Coding (PSC) for image retrieval. In our framework, multiple global and local features are extracted from images and then are transformed to Trimmed-Root (TR)-features. After that, the features are encoded into compact codes by PSC. Finally, a two-stage ranking strategy is proposed for indexing in retrieval. We make three major contributions in this study. First, we propose TR representation of multiple image features and show that the TR representation offers better performance than the original features. Second, the integrated features by PSC is very compact and effective with lower complexity than by the standard sparse coding. Finally, the two-stage ranking strategy can balance the efficiency and memory usage in storage. Experiments demonstrate that our compact image representation is superior to the state-of-the-art alternatives for large-scale image retrieval.

AB - In this study, we propose a simple, yet general and powerful framework of integrating multiple global and local features by Product Sparse Coding (PSC) for image retrieval. In our framework, multiple global and local features are extracted from images and then are transformed to Trimmed-Root (TR)-features. After that, the features are encoded into compact codes by PSC. Finally, a two-stage ranking strategy is proposed for indexing in retrieval. We make three major contributions in this study. First, we propose TR representation of multiple image features and show that the TR representation offers better performance than the original features. Second, the integrated features by PSC is very compact and effective with lower complexity than by the standard sparse coding. Finally, the two-stage ranking strategy can balance the efficiency and memory usage in storage. Experiments demonstrate that our compact image representation is superior to the state-of-the-art alternatives for large-scale image retrieval.

KW - Image representation

KW - Image retrieval

KW - Product Sparse Coding (PSC)

KW - Ranking strategy

KW - Trimmed-Root (TR)-feature

UR - http://www.scopus.com/inward/record.url?scp=84959497522&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959497522&partnerID=8YFLogxK

U2 - 10.1587/transinf.2015EDP7337

DO - 10.1587/transinf.2015EDP7337

M3 - Article

VL - E99D

SP - 731

EP - 738

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 3

ER -