Multi-class relevance feedback content-based image retrieval

Research output: Contribution to journalArticle

21 Citations (Scopus)

Abstract

Relevance feedback methods for content-based image retrieval have shown promise in a variety of image database applications. These techniques assume two-class relevance feedback: relevant and irrelevant classes. While simple computationally, two-class relevance feedback often becomes inadequate in providing sufficient information to help rapidly improve retrieval performance. In this paper we propose a multi-class form of relevance feedback retrieval to try to exploit multi-class information. For a given query, we use a Χ 2 analysis to determine the local relevance of each feature dimension with multi-class relevance feedback. This information is then used to customize the retrieval metric to rank images. By exploiting multi-class information, our method is able to create flexible metrics that better capture user perceived similarity. In a number of image data sets, the method achieves a higher level of precision with fewer iterations, demonstrating the potential for substantial improvements over two-class relevance feedback retrieval.

Original languageEnglish
Pages (from-to)42-67
Number of pages26
JournalComputer Vision and Image Understanding
Volume90
Issue number1
DOIs
StatePublished - 1 Jan 2003

Fingerprint

Image retrieval
Feedback

Cite this

@article{4810d1d55f214ec7a27f6996b83e56a8,
title = "Multi-class relevance feedback content-based image retrieval",
abstract = "Relevance feedback methods for content-based image retrieval have shown promise in a variety of image database applications. These techniques assume two-class relevance feedback: relevant and irrelevant classes. While simple computationally, two-class relevance feedback often becomes inadequate in providing sufficient information to help rapidly improve retrieval performance. In this paper we propose a multi-class form of relevance feedback retrieval to try to exploit multi-class information. For a given query, we use a Χ 2 analysis to determine the local relevance of each feature dimension with multi-class relevance feedback. This information is then used to customize the retrieval metric to rank images. By exploiting multi-class information, our method is able to create flexible metrics that better capture user perceived similarity. In a number of image data sets, the method achieves a higher level of precision with fewer iterations, demonstrating the potential for substantial improvements over two-class relevance feedback retrieval.",
author = "Jing Peng",
year = "2003",
month = "1",
day = "1",
doi = "10.1016/S1077-3142(03)00013-4",
language = "English",
volume = "90",
pages = "42--67",
journal = "Computer Vision and Image Understanding",
issn = "1077-3142",
publisher = "Academic Press Inc.",
number = "1",

}

Multi-class relevance feedback content-based image retrieval. / Peng, Jing.

In: Computer Vision and Image Understanding, Vol. 90, No. 1, 01.01.2003, p. 42-67.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Multi-class relevance feedback content-based image retrieval

AU - Peng, Jing

PY - 2003/1/1

Y1 - 2003/1/1

N2 - Relevance feedback methods for content-based image retrieval have shown promise in a variety of image database applications. These techniques assume two-class relevance feedback: relevant and irrelevant classes. While simple computationally, two-class relevance feedback often becomes inadequate in providing sufficient information to help rapidly improve retrieval performance. In this paper we propose a multi-class form of relevance feedback retrieval to try to exploit multi-class information. For a given query, we use a Χ 2 analysis to determine the local relevance of each feature dimension with multi-class relevance feedback. This information is then used to customize the retrieval metric to rank images. By exploiting multi-class information, our method is able to create flexible metrics that better capture user perceived similarity. In a number of image data sets, the method achieves a higher level of precision with fewer iterations, demonstrating the potential for substantial improvements over two-class relevance feedback retrieval.

AB - Relevance feedback methods for content-based image retrieval have shown promise in a variety of image database applications. These techniques assume two-class relevance feedback: relevant and irrelevant classes. While simple computationally, two-class relevance feedback often becomes inadequate in providing sufficient information to help rapidly improve retrieval performance. In this paper we propose a multi-class form of relevance feedback retrieval to try to exploit multi-class information. For a given query, we use a Χ 2 analysis to determine the local relevance of each feature dimension with multi-class relevance feedback. This information is then used to customize the retrieval metric to rank images. By exploiting multi-class information, our method is able to create flexible metrics that better capture user perceived similarity. In a number of image data sets, the method achieves a higher level of precision with fewer iterations, demonstrating the potential for substantial improvements over two-class relevance feedback retrieval.

UR - http://www.scopus.com/inward/record.url?scp=0242573586&partnerID=8YFLogxK

U2 - 10.1016/S1077-3142(03)00013-4

DO - 10.1016/S1077-3142(03)00013-4

M3 - Article

AN - SCOPUS:0242573586

VL - 90

SP - 42

EP - 67

JO - Computer Vision and Image Understanding

JF - Computer Vision and Image Understanding

SN - 1077-3142

IS - 1

ER -