Incorporating real-valued multiple instance learning into relevance feedback for image retrieval

Xin Hunag, Shu Ching Chen, Mei-Ling Shyu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

10 Citations (Scopus)

Abstract

This paper presents a content-based image retrieval (CBIR) system that incorporates real-valued multiple instance learning (MIL) into the user relevance feedback (RF) to learn the user's subjective visual concepts, especially where the user's most interested region and how to map the local feature vector of that region to the high-level concept pattern of the user. RF provides a way to obtain the subjectivity of the user's high-level visual concepts, and MIL enables the automatic learning of the user's high-level concepts. The user interacts with the CBIR system by relevance feedback in a way that the extent to which the image samples retrieved by the system are relevant to the user's intention is labeled. The system in turn applies the MIL method to find user's most interested image region from the feedback. A multilayer neural network that is trained progressively through the feedback and learning procedure is used to map the low-level image features to the high-level concepts.

Original languageEnglish
Title of host publicationProceedings - IEEE International Conference on Multimedia and Expo
PublisherIEEE Computer Society
PagesI321-I324
Volume1
ISBN (Print)0780379659
DOIs
StatePublished - Jan 1 2003
Event2003 International Conference on Multimedia and Expo, ICME 2003 - Baltimore, United States
Duration: Jul 6 2003Jul 9 2003

Other

Other2003 International Conference on Multimedia and Expo, ICME 2003
CountryUnited States
CityBaltimore
Period7/6/037/9/03

Fingerprint

Image retrieval
Feedback
Multilayer neural networks

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Computer Science Applications

Cite this

Hunag, X., Chen, S. C., & Shyu, M-L. (2003). Incorporating real-valued multiple instance learning into relevance feedback for image retrieval. In Proceedings - IEEE International Conference on Multimedia and Expo (Vol. 1, pp. I321-I324). [1220919] IEEE Computer Society. https://doi.org/10.1109/ICME.2003.1220919

Incorporating real-valued multiple instance learning into relevance feedback for image retrieval. / Hunag, Xin; Chen, Shu Ching; Shyu, Mei-Ling.

Proceedings - IEEE International Conference on Multimedia and Expo. Vol. 1 IEEE Computer Society, 2003. p. I321-I324 1220919.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Hunag, X, Chen, SC & Shyu, M-L 2003, Incorporating real-valued multiple instance learning into relevance feedback for image retrieval. in Proceedings - IEEE International Conference on Multimedia and Expo. vol. 1, 1220919, IEEE Computer Society, pp. I321-I324, 2003 International Conference on Multimedia and Expo, ICME 2003, Baltimore, United States, 7/6/03. https://doi.org/10.1109/ICME.2003.1220919
Hunag X, Chen SC, Shyu M-L. Incorporating real-valued multiple instance learning into relevance feedback for image retrieval. In Proceedings - IEEE International Conference on Multimedia and Expo. Vol. 1. IEEE Computer Society. 2003. p. I321-I324. 1220919 https://doi.org/10.1109/ICME.2003.1220919
Hunag, Xin ; Chen, Shu Ching ; Shyu, Mei-Ling. / Incorporating real-valued multiple instance learning into relevance feedback for image retrieval. Proceedings - IEEE International Conference on Multimedia and Expo. Vol. 1 IEEE Computer Society, 2003. pp. I321-I324
@inproceedings{cde13d124f144db89bdef387f827ffbf,
title = "Incorporating real-valued multiple instance learning into relevance feedback for image retrieval",
abstract = "This paper presents a content-based image retrieval (CBIR) system that incorporates real-valued multiple instance learning (MIL) into the user relevance feedback (RF) to learn the user's subjective visual concepts, especially where the user's most interested region and how to map the local feature vector of that region to the high-level concept pattern of the user. RF provides a way to obtain the subjectivity of the user's high-level visual concepts, and MIL enables the automatic learning of the user's high-level concepts. The user interacts with the CBIR system by relevance feedback in a way that the extent to which the image samples retrieved by the system are relevant to the user's intention is labeled. The system in turn applies the MIL method to find user's most interested image region from the feedback. A multilayer neural network that is trained progressively through the feedback and learning procedure is used to map the low-level image features to the high-level concepts.",
author = "Xin Hunag and Chen, {Shu Ching} and Mei-Ling Shyu",
year = "2003",
month = "1",
day = "1",
doi = "10.1109/ICME.2003.1220919",
language = "English",
isbn = "0780379659",
volume = "1",
pages = "I321--I324",
booktitle = "Proceedings - IEEE International Conference on Multimedia and Expo",
publisher = "IEEE Computer Society",

}

TY - GEN

T1 - Incorporating real-valued multiple instance learning into relevance feedback for image retrieval

AU - Hunag, Xin

AU - Chen, Shu Ching

AU - Shyu, Mei-Ling

PY - 2003/1/1

Y1 - 2003/1/1

N2 - This paper presents a content-based image retrieval (CBIR) system that incorporates real-valued multiple instance learning (MIL) into the user relevance feedback (RF) to learn the user's subjective visual concepts, especially where the user's most interested region and how to map the local feature vector of that region to the high-level concept pattern of the user. RF provides a way to obtain the subjectivity of the user's high-level visual concepts, and MIL enables the automatic learning of the user's high-level concepts. The user interacts with the CBIR system by relevance feedback in a way that the extent to which the image samples retrieved by the system are relevant to the user's intention is labeled. The system in turn applies the MIL method to find user's most interested image region from the feedback. A multilayer neural network that is trained progressively through the feedback and learning procedure is used to map the low-level image features to the high-level concepts.

AB - This paper presents a content-based image retrieval (CBIR) system that incorporates real-valued multiple instance learning (MIL) into the user relevance feedback (RF) to learn the user's subjective visual concepts, especially where the user's most interested region and how to map the local feature vector of that region to the high-level concept pattern of the user. RF provides a way to obtain the subjectivity of the user's high-level visual concepts, and MIL enables the automatic learning of the user's high-level concepts. The user interacts with the CBIR system by relevance feedback in a way that the extent to which the image samples retrieved by the system are relevant to the user's intention is labeled. The system in turn applies the MIL method to find user's most interested image region from the feedback. A multilayer neural network that is trained progressively through the feedback and learning procedure is used to map the low-level image features to the high-level concepts.

UR - http://www.scopus.com/inward/record.url?scp=84897804159&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84897804159&partnerID=8YFLogxK

U2 - 10.1109/ICME.2003.1220919

DO - 10.1109/ICME.2003.1220919

M3 - Conference contribution

SN - 0780379659

VL - 1

SP - I321-I324

BT - Proceedings - IEEE International Conference on Multimedia and Expo

PB - IEEE Computer Society

ER -