Supervised discriminative hashing for compact binary codes

Viet Anh Nguyen, Jiwen Lu, Minh N Do

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

Abstract

Binary hashing has been increasingly popular for efficient similarity search in large-scale vision problems. This paper presents a novel Supervised Discriminative Hashing (SDH) method by jointly modeling the global and local manifold structures. Specifically, a family of discriminative hash functions is designed to map data points of the original highdimensional space into nearby compact binary codes while preserving the geometrical similarity and discriminant properties in both global and local neighborhoods. Furthermore, the quantization loss between the original data and the binary codes together with the even binary code distribution are also taken into account in the optimization to generate more efficient and compact binary codes. Experimental results have demonstrated the proposed method outperforms the state-of-the-art.

Original languageEnglish (US)
Title of host publicationMM 2014 - Proceedings of the 2014 ACM Conference on Multimedia
PublisherAssociation for Computing Machinery, Inc
Pages989-992
Number of pages4
ISBN (Electronic)9781450330633
DOIs
StatePublished - Nov 3 2014
Event2014 ACM Conference on Multimedia, MM 2014 - Orlando, United States
Duration: Nov 3 2014Nov 7 2014

Publication series

NameMM 2014 - Proceedings of the 2014 ACM Conference on Multimedia

Other

Other2014 ACM Conference on Multimedia, MM 2014
CountryUnited States
CityOrlando
Period11/3/1411/7/14

Fingerprint

Binary codes
Hash functions

Keywords

  • Approximate nearest neighbor search
  • Binary codes
  • Hashing
  • Supervised hashing

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition
  • Media Technology
  • Software

Cite this

Nguyen, V. A., Lu, J., & Do, M. N. (2014). Supervised discriminative hashing for compact binary codes. In MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia (pp. 989-992). (MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia). Association for Computing Machinery, Inc. https://doi.org/10.1145/2647868.2655003

Supervised discriminative hashing for compact binary codes. / Nguyen, Viet Anh; Lu, Jiwen; Do, Minh N.

MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia. Association for Computing Machinery, Inc, 2014. p. 989-992 (MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia).

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

Nguyen, VA, Lu, J & Do, MN 2014, Supervised discriminative hashing for compact binary codes. in MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia. MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia, Association for Computing Machinery, Inc, pp. 989-992, 2014 ACM Conference on Multimedia, MM 2014, Orlando, United States, 11/3/14. https://doi.org/10.1145/2647868.2655003
Nguyen VA, Lu J, Do MN. Supervised discriminative hashing for compact binary codes. In MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia. Association for Computing Machinery, Inc. 2014. p. 989-992. (MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia). https://doi.org/10.1145/2647868.2655003
Nguyen, Viet Anh ; Lu, Jiwen ; Do, Minh N. / Supervised discriminative hashing for compact binary codes. MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia. Association for Computing Machinery, Inc, 2014. pp. 989-992 (MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia).
@inproceedings{76b2ccfa22074e6380721ac9bf16b21e,
title = "Supervised discriminative hashing for compact binary codes",
abstract = "Binary hashing has been increasingly popular for efficient similarity search in large-scale vision problems. This paper presents a novel Supervised Discriminative Hashing (SDH) method by jointly modeling the global and local manifold structures. Specifically, a family of discriminative hash functions is designed to map data points of the original highdimensional space into nearby compact binary codes while preserving the geometrical similarity and discriminant properties in both global and local neighborhoods. Furthermore, the quantization loss between the original data and the binary codes together with the even binary code distribution are also taken into account in the optimization to generate more efficient and compact binary codes. Experimental results have demonstrated the proposed method outperforms the state-of-the-art.",
keywords = "Approximate nearest neighbor search, Binary codes, Hashing, Supervised hashing",
author = "Nguyen, {Viet Anh} and Jiwen Lu and Do, {Minh N}",
year = "2014",
month = "11",
day = "3",
doi = "10.1145/2647868.2655003",
language = "English (US)",
series = "MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia",
publisher = "Association for Computing Machinery, Inc",
pages = "989--992",
booktitle = "MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia",

}

TY - GEN

T1 - Supervised discriminative hashing for compact binary codes

AU - Nguyen, Viet Anh

AU - Lu, Jiwen

AU - Do, Minh N

PY - 2014/11/3

Y1 - 2014/11/3

N2 - Binary hashing has been increasingly popular for efficient similarity search in large-scale vision problems. This paper presents a novel Supervised Discriminative Hashing (SDH) method by jointly modeling the global and local manifold structures. Specifically, a family of discriminative hash functions is designed to map data points of the original highdimensional space into nearby compact binary codes while preserving the geometrical similarity and discriminant properties in both global and local neighborhoods. Furthermore, the quantization loss between the original data and the binary codes together with the even binary code distribution are also taken into account in the optimization to generate more efficient and compact binary codes. Experimental results have demonstrated the proposed method outperforms the state-of-the-art.

AB - Binary hashing has been increasingly popular for efficient similarity search in large-scale vision problems. This paper presents a novel Supervised Discriminative Hashing (SDH) method by jointly modeling the global and local manifold structures. Specifically, a family of discriminative hash functions is designed to map data points of the original highdimensional space into nearby compact binary codes while preserving the geometrical similarity and discriminant properties in both global and local neighborhoods. Furthermore, the quantization loss between the original data and the binary codes together with the even binary code distribution are also taken into account in the optimization to generate more efficient and compact binary codes. Experimental results have demonstrated the proposed method outperforms the state-of-the-art.

KW - Approximate nearest neighbor search

KW - Binary codes

KW - Hashing

KW - Supervised hashing

UR - http://www.scopus.com/inward/record.url?scp=84913534887&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84913534887&partnerID=8YFLogxK

U2 - 10.1145/2647868.2655003

DO - 10.1145/2647868.2655003

M3 - Conference contribution

T3 - MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia

SP - 989

EP - 992

BT - MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia

PB - Association for Computing Machinery, Inc

ER -