Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label

Paul H Kwan; Keisuke Kameyama; Kazuo Toraichi

doi:10.1109/FUZZ.2001.1008916

Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label

Paul H Kwan, Keisuke Kameyama, Kazuo Toraichi

Computer Science

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

3 Citations (Scopus)

Abstract

Our work in image similarity retrieval by relaxation labeling processes required adding local constraints to obtain an initial set of compatible objects and labels on pairs of images to ensure labeling consistency at convergence. This approach suffers from the problem of over-specified constraints that leads to potentially invaluable information being prematurely removed. To address this problem, we introduce the idea of a Match-all label for objects that failed these constraints. It serves to give them a defined labeling probability as well as allowing them participate in global correspondences through the compatibility model. We show that this enhanced formulation still meets the conditions for a theorem on labeling consistency in Hummel and Zucker to be satisfied. At convergence, the set of objects and their most consistent labels constitute a best partial labeling for the pair of images.

Original language	English
Title of host publication	Meeting the grand challenge: machines that serve people - Proceedings of the 10th IEEE International Conference on Fuzzy Systems
Place of Publication	Los Alamitos, United States of America
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Pages	1384-1387
Volume	3
ISBN (Print)	078037293X
DOIs	https://doi.org/10.1109/FUZZ.2001.1008916
Publication status	Published - 2001
Event	FUZZ-IEEE 2001: 10th IEEE International Conference on Fuzzy Systems - Melbourne, Australia Duration: 2 Dec 2001 → 5 Dec 2001

Conference

Conference	FUZZ-IEEE 2001: 10th IEEE International Conference on Fuzzy Systems
City	Melbourne, Australia
Period	2/12/01 → 5/12/01

Keywords

Neural, Evolutionary and Fuzzy Computation
Image Processing

Access to Document

10.1109/FUZZ.2001.1008916

https://hdl.handle.net/1959.11/5989

Handle.net

Persistent link

Cite this

Kwan, P. H., Kameyama, K., & Toraichi, K. (2001). Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label. In Meeting the grand challenge: machines that serve people - Proceedings of the 10th IEEE International Conference on Fuzzy Systems (Vol. 3, pp. 1384-1387). Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.1109/FUZZ.2001.1008916

Kwan, Paul H ; Kameyama, Keisuke ; Toraichi, Kazuo. / Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label. Meeting the grand challenge: machines that serve people - Proceedings of the 10th IEEE International Conference on Fuzzy Systems. Vol. 3 Los Alamitos, United States of America : Institute of Electrical and Electronics Engineers (IEEE), 2001. pp. 1384-1387

@inproceedings{f4bb0ad8e51f4c69a6d37b100f0fd29f,

title = "Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label",

abstract = "Our work in image similarity retrieval by relaxation labeling processes required adding local constraints to obtain an initial set of compatible objects and labels on pairs of images to ensure labeling consistency at convergence. This approach suffers from the problem of over-specified constraints that leads to potentially invaluable information being prematurely removed. To address this problem, we introduce the idea of a Match-all label for objects that failed these constraints. It serves to give them a defined labeling probability as well as allowing them participate in global correspondences through the compatibility model. We show that this enhanced formulation still meets the conditions for a theorem on labeling consistency in Hummel and Zucker to be satisfied. At convergence, the set of objects and their most consistent labels constitute a best partial labeling for the pair of images.",

keywords = "Neural, Evolutionary and Fuzzy Computation, Image Processing",

author = "Kwan, \{Paul H\} and Keisuke Kameyama and Kazuo Toraichi",

year = "2001",

doi = "10.1109/FUZZ.2001.1008916",

language = "English",

isbn = "078037293X",

volume = "3",

pages = "1384--1387",

booktitle = "Meeting the grand challenge: machines that serve people - Proceedings of the 10th IEEE International Conference on Fuzzy Systems",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

note = "FUZZ-IEEE 2001: 10th IEEE International Conference on Fuzzy Systems ; Conference date: 02-12-2001 Through 05-12-2001",

}

Kwan, PH, Kameyama, K & Toraichi, K 2001, Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label. in Meeting the grand challenge: machines that serve people - Proceedings of the 10th IEEE International Conference on Fuzzy Systems. vol. 3, Institute of Electrical and Electronics Engineers (IEEE), Los Alamitos, United States of America, pp. 1384-1387, FUZZ-IEEE 2001: 10th IEEE International Conference on Fuzzy Systems, Melbourne, Australia, 2/12/01. https://doi.org/10.1109/FUZZ.2001.1008916

Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label. / Kwan, Paul H; Kameyama, Keisuke; Toraichi, Kazuo.
Meeting the grand challenge: machines that serve people - Proceedings of the 10th IEEE International Conference on Fuzzy Systems. Vol. 3 Los Alamitos, United States of America: Institute of Electrical and Electronics Engineers (IEEE), 2001. p. 1384-1387.

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

TY - GEN

T1 - Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label

AU - Kwan, Paul H

AU - Kameyama, Keisuke

AU - Toraichi, Kazuo

PY - 2001

Y1 - 2001

N2 - Our work in image similarity retrieval by relaxation labeling processes required adding local constraints to obtain an initial set of compatible objects and labels on pairs of images to ensure labeling consistency at convergence. This approach suffers from the problem of over-specified constraints that leads to potentially invaluable information being prematurely removed. To address this problem, we introduce the idea of a Match-all label for objects that failed these constraints. It serves to give them a defined labeling probability as well as allowing them participate in global correspondences through the compatibility model. We show that this enhanced formulation still meets the conditions for a theorem on labeling consistency in Hummel and Zucker to be satisfied. At convergence, the set of objects and their most consistent labels constitute a best partial labeling for the pair of images.

AB - Our work in image similarity retrieval by relaxation labeling processes required adding local constraints to obtain an initial set of compatible objects and labels on pairs of images to ensure labeling consistency at convergence. This approach suffers from the problem of over-specified constraints that leads to potentially invaluable information being prematurely removed. To address this problem, we introduce the idea of a Match-all label for objects that failed these constraints. It serves to give them a defined labeling probability as well as allowing them participate in global correspondences through the compatibility model. We show that this enhanced formulation still meets the conditions for a theorem on labeling consistency in Hummel and Zucker to be satisfied. At convergence, the set of objects and their most consistent labels constitute a best partial labeling for the pair of images.

KW - Neural, Evolutionary and Fuzzy Computation

KW - Image Processing

UR - http://trove.nla.gov.au/work/16793499

UR - https://www.scopus.com/pages/publications/0036343349

U2 - 10.1109/FUZZ.2001.1008916

DO - 10.1109/FUZZ.2001.1008916

M3 - Conference contribution

SN - 078037293X

VL - 3

SP - 1384

EP - 1387

BT - Meeting the grand challenge: machines that serve people - Proceedings of the 10th IEEE International Conference on Fuzzy Systems

PB - Institute of Electrical and Electronics Engineers (IEEE)

CY - Los Alamitos, United States of America

T2 - FUZZ-IEEE 2001: 10th IEEE International Conference on Fuzzy Systems

Y2 - 2 December 2001 through 5 December 2001

ER -

Kwan PH, Kameyama K, Toraichi K. Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label. In Meeting the grand challenge: machines that serve people - Proceedings of the 10th IEEE International Conference on Fuzzy Systems. Vol. 3. Los Alamitos, United States of America: Institute of Electrical and Electronics Engineers (IEEE). 2001. p. 1384-1387 doi: 10.1109/FUZZ.2001.1008916

Connecting Image Similarity Retrieval with Consistent Labeling Problem by Introducing a Match-all Label

Abstract

Conference

Keywords

Access to Document

Handle.net

Fingerprint

Cite this