Abstract

Discriminative dictionary learning has been widely used in many applications such as face retrieval / recognition and image classification, where the labels of the training data are utilized to improve the discriminative power of the learned dictionary. This paper deals with a new problem of learning a dictionary for associating pairs of images in applications such as face image retrieval. Compared with a typical supervised learning task, in this case the labeling information is very limited (e.g. only some training pairs are known to be associated). Further, associated pairs may be considered similar only after excluding certain regions (e.g. sunglasses in a face image). We formulate a dictionary learning problem under these considerations and design an algorithm to solve the problem. We also provide a proof for the convergence of the algorithm. Experiments and results suggest that the proposed method is advantageous over common baselines.

Original languageEnglish (US)
Title of host publication2017 IEEE International Conference on Multimedia and Expo, ICME 2017
PublisherIEEE Computer Society
Pages1410-1415
Number of pages6
ISBN (Electronic)9781509060672
DOIs
StatePublished - Aug 28 2017
Event2017 IEEE International Conference on Multimedia and Expo, ICME 2017 - Hong Kong, Hong Kong
Duration: Jul 10 2017Jul 14 2017

Other

Other2017 IEEE International Conference on Multimedia and Expo, ICME 2017
CountryHong Kong
CityHong Kong
Period7/10/177/14/17

Fingerprint

Glossaries
Image classification
Supervised learning
Image retrieval
Labeling
Labels
Experiments

Keywords

  • Constrained dictionary learning
  • Face retrieval
  • Non-negative

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Computer Science Applications

Cite this

Zhou, X., Ding, P. L. K., & Li, B. (2017). Non-negative dictionary learning with pairwise partial similarity constraint. In 2017 IEEE International Conference on Multimedia and Expo, ICME 2017 (pp. 1410-1415). [8019392] IEEE Computer Society. https://doi.org/10.1109/ICME.2017.8019392

Non-negative dictionary learning with pairwise partial similarity constraint. / Zhou, Xu; Ding, Pak Lun Kevin; Li, Baoxin.

2017 IEEE International Conference on Multimedia and Expo, ICME 2017. IEEE Computer Society, 2017. p. 1410-1415 8019392.

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

Zhou, X, Ding, PLK & Li, B 2017, Non-negative dictionary learning with pairwise partial similarity constraint. in 2017 IEEE International Conference on Multimedia and Expo, ICME 2017., 8019392, IEEE Computer Society, pp. 1410-1415, 2017 IEEE International Conference on Multimedia and Expo, ICME 2017, Hong Kong, Hong Kong, 7/10/17. https://doi.org/10.1109/ICME.2017.8019392
Zhou X, Ding PLK, Li B. Non-negative dictionary learning with pairwise partial similarity constraint. In 2017 IEEE International Conference on Multimedia and Expo, ICME 2017. IEEE Computer Society. 2017. p. 1410-1415. 8019392 https://doi.org/10.1109/ICME.2017.8019392
Zhou, Xu ; Ding, Pak Lun Kevin ; Li, Baoxin. / Non-negative dictionary learning with pairwise partial similarity constraint. 2017 IEEE International Conference on Multimedia and Expo, ICME 2017. IEEE Computer Society, 2017. pp. 1410-1415
@inproceedings{41cd49bbccd349538441ac973e53b023,
title = "Non-negative dictionary learning with pairwise partial similarity constraint",
abstract = "Discriminative dictionary learning has been widely used in many applications such as face retrieval / recognition and image classification, where the labels of the training data are utilized to improve the discriminative power of the learned dictionary. This paper deals with a new problem of learning a dictionary for associating pairs of images in applications such as face image retrieval. Compared with a typical supervised learning task, in this case the labeling information is very limited (e.g. only some training pairs are known to be associated). Further, associated pairs may be considered similar only after excluding certain regions (e.g. sunglasses in a face image). We formulate a dictionary learning problem under these considerations and design an algorithm to solve the problem. We also provide a proof for the convergence of the algorithm. Experiments and results suggest that the proposed method is advantageous over common baselines.",
keywords = "Constrained dictionary learning, Face retrieval, Non-negative",
author = "Xu Zhou and Ding, {Pak Lun Kevin} and Baoxin Li",
year = "2017",
month = "8",
day = "28",
doi = "10.1109/ICME.2017.8019392",
language = "English (US)",
pages = "1410--1415",
booktitle = "2017 IEEE International Conference on Multimedia and Expo, ICME 2017",
publisher = "IEEE Computer Society",
address = "United States",

}

TY - GEN

T1 - Non-negative dictionary learning with pairwise partial similarity constraint

AU - Zhou, Xu

AU - Ding, Pak Lun Kevin

AU - Li, Baoxin

PY - 2017/8/28

Y1 - 2017/8/28

N2 - Discriminative dictionary learning has been widely used in many applications such as face retrieval / recognition and image classification, where the labels of the training data are utilized to improve the discriminative power of the learned dictionary. This paper deals with a new problem of learning a dictionary for associating pairs of images in applications such as face image retrieval. Compared with a typical supervised learning task, in this case the labeling information is very limited (e.g. only some training pairs are known to be associated). Further, associated pairs may be considered similar only after excluding certain regions (e.g. sunglasses in a face image). We formulate a dictionary learning problem under these considerations and design an algorithm to solve the problem. We also provide a proof for the convergence of the algorithm. Experiments and results suggest that the proposed method is advantageous over common baselines.

AB - Discriminative dictionary learning has been widely used in many applications such as face retrieval / recognition and image classification, where the labels of the training data are utilized to improve the discriminative power of the learned dictionary. This paper deals with a new problem of learning a dictionary for associating pairs of images in applications such as face image retrieval. Compared with a typical supervised learning task, in this case the labeling information is very limited (e.g. only some training pairs are known to be associated). Further, associated pairs may be considered similar only after excluding certain regions (e.g. sunglasses in a face image). We formulate a dictionary learning problem under these considerations and design an algorithm to solve the problem. We also provide a proof for the convergence of the algorithm. Experiments and results suggest that the proposed method is advantageous over common baselines.

KW - Constrained dictionary learning

KW - Face retrieval

KW - Non-negative

UR - http://www.scopus.com/inward/record.url?scp=85030246689&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85030246689&partnerID=8YFLogxK

U2 - 10.1109/ICME.2017.8019392

DO - 10.1109/ICME.2017.8019392

M3 - Conference contribution

SP - 1410

EP - 1415

BT - 2017 IEEE International Conference on Multimedia and Expo, ICME 2017

PB - IEEE Computer Society

ER -