Localized content based image retrieval by multiple instance active learning

Dan Zhang; Fei Wang; Zhenwei Shi; Changshui Zhang

doi:10.1109/ICIP.2008.4711906

Localized content based image retrieval by multiple instance active learning

Dan Zhang^*
, Fei Wang
, Zhenwei Shi
, Changshui Zhang

^*Corresponding author for this work

School of Astronautics

Tsinghua University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

3 Scopus citations

Abstract

In this paper, we propose two general multiple instance active learning (MIAL) algorithms, multiple-instance active learning with a simple margin strategy (S-MIAL) and multiple-instance active learning with fisher information (F-MIAL), and apply them to the relevance feedback in localized content based image retrieval (LCBIR). S-MIAL considers the most ambiguous picture as the most valuable one, while F-MIAL can utilize the fisher information and analyze the value of the unlabeled pictures by assigning different labels to them. We show that F-MIAL can be integrated more naturally into the multiple instance learning scenario. In experiments, we will show their superior performances on some real-world image datasets.

Original language	English
Title of host publication	2008 IEEE International Conference on Image Processing, ICIP 2008 Proceedings
Pages	921-924
Number of pages	4
DOIs	https://doi.org/10.1109/ICIP.2008.4711906
State	Published - 2008
Event	2008 IEEE International Conference on Image Processing, ICIP 2008 - San Diego, CA, United States Duration: 12 Oct 2008 → 15 Oct 2008

Publication series

Name	Proceedings - International Conference on Image Processing, ICIP
ISSN (Print)	1522-4880

Conference

Conference	2008 IEEE International Conference on Image Processing, ICIP 2008
Country/Territory	United States
City	San Diego, CA
Period	12/10/08 → 15/10/08

Keywords

Fisher information
Localized content based image retrieval
Multiple instance active learning
Relevance feedback

Access to Document

10.1109/ICIP.2008.4711906

Cite this

@inproceedings{57e6f47d209d4aceb97d2c35e7168d0b,

title = "Localized content based image retrieval by multiple instance active learning",

abstract = "In this paper, we propose two general multiple instance active learning (MIAL) algorithms, multiple-instance active learning with a simple margin strategy (S-MIAL) and multiple-instance active learning with fisher information (F-MIAL), and apply them to the relevance feedback in localized content based image retrieval (LCBIR). S-MIAL considers the most ambiguous picture as the most valuable one, while F-MIAL can utilize the fisher information and analyze the value of the unlabeled pictures by assigning different labels to them. We show that F-MIAL can be integrated more naturally into the multiple instance learning scenario. In experiments, we will show their superior performances on some real-world image datasets.",

keywords = "Fisher information, Localized content based image retrieval, Multiple instance active learning, Relevance feedback",

author = "Dan Zhang and Fei Wang and Zhenwei Shi and Changshui Zhang",

year = "2008",

doi = "10.1109/ICIP.2008.4711906",

language = "英语",

isbn = "1424417643",

series = "Proceedings - International Conference on Image Processing, ICIP",

pages = "921--924",

booktitle = "2008 IEEE International Conference on Image Processing, ICIP 2008 Proceedings",

note = "2008 IEEE International Conference on Image Processing, ICIP 2008 ; Conference date: 12-10-2008 Through 15-10-2008",

}

Zhang, D, Wang, F, Shi, Z & Zhang, C 2008, Localized content based image retrieval by multiple instance active learning. in 2008 IEEE International Conference on Image Processing, ICIP 2008 Proceedings., 4711906, Proceedings - International Conference on Image Processing, ICIP, pp. 921-924, 2008 IEEE International Conference on Image Processing, ICIP 2008, San Diego, CA, United States, 12/10/08. https://doi.org/10.1109/ICIP.2008.4711906

Localized content based image retrieval by multiple instance active learning. / Zhang, Dan; Wang, Fei; Shi, Zhenwei et al.
2008 IEEE International Conference on Image Processing, ICIP 2008 Proceedings. 2008. p. 921-924 4711906 (Proceedings - International Conference on Image Processing, ICIP).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Localized content based image retrieval by multiple instance active learning

AU - Zhang, Dan

AU - Wang, Fei

AU - Shi, Zhenwei

AU - Zhang, Changshui

PY - 2008

Y1 - 2008

N2 - In this paper, we propose two general multiple instance active learning (MIAL) algorithms, multiple-instance active learning with a simple margin strategy (S-MIAL) and multiple-instance active learning with fisher information (F-MIAL), and apply them to the relevance feedback in localized content based image retrieval (LCBIR). S-MIAL considers the most ambiguous picture as the most valuable one, while F-MIAL can utilize the fisher information and analyze the value of the unlabeled pictures by assigning different labels to them. We show that F-MIAL can be integrated more naturally into the multiple instance learning scenario. In experiments, we will show their superior performances on some real-world image datasets.

AB - In this paper, we propose two general multiple instance active learning (MIAL) algorithms, multiple-instance active learning with a simple margin strategy (S-MIAL) and multiple-instance active learning with fisher information (F-MIAL), and apply them to the relevance feedback in localized content based image retrieval (LCBIR). S-MIAL considers the most ambiguous picture as the most valuable one, while F-MIAL can utilize the fisher information and analyze the value of the unlabeled pictures by assigning different labels to them. We show that F-MIAL can be integrated more naturally into the multiple instance learning scenario. In experiments, we will show their superior performances on some real-world image datasets.

KW - Fisher information

KW - Localized content based image retrieval

KW - Multiple instance active learning

KW - Relevance feedback

UR - https://www.scopus.com/pages/publications/69949122999

U2 - 10.1109/ICIP.2008.4711906

DO - 10.1109/ICIP.2008.4711906

M3 - 会议稿件

AN - SCOPUS:69949122999

SN - 1424417643

SN - 9781424417643

T3 - Proceedings - International Conference on Image Processing, ICIP

SP - 921

EP - 924

BT - 2008 IEEE International Conference on Image Processing, ICIP 2008 Proceedings

T2 - 2008 IEEE International Conference on Image Processing, ICIP 2008

Y2 - 12 October 2008 through 15 October 2008

ER -

Localized content based image retrieval by multiple instance active learning

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this