Rank Constraints on Joint Dictionary Learning for Image Recognition

DOI

10.2991/icmmcce-15.2015.538How to use a DOI?

Keywords

Rank constraints, Joint dictionary learning, Image recognition.

Abstract

Sparse representation has been extensively applied to image recognition, while learning appropriate dictionaries for image content representation plays a critical role in it. An approach to simultaneously learn a common dictionary and multiple class-specific particular dictionaries achieves state-of-the-art performance. However, how to separate the particularity and commonality correctly is a quite important problem. Meanwhile, there exists an over-fitting phenomenon in the dictionary learned from given training samples which lie on a low dimensional subspace in that data samples can be linearly represented by dictionary. In this paper, we propose rank constraints on the joint dictionary learning (RC-JDL) algorithm to solve the above questions. Extensive experimental results on public available databases demonstrate the effectiveness of the proposed algorithm.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Haohao Meng
AU  - Yufeng Chen
PY  - 2015/12
DA  - 2015/12
TI  - Rank Constraints on Joint Dictionary Learning for Image Recognition
BT  - Proceedings of the 4th International Conference on Mechatronics, Materials, Chemistry and Computer Engineering 2015
PB  - Atlantis Press
SN  - 2352-538X
UR  - https://doi.org/10.2991/icmmcce-15.2015.538
DO  - 10.2991/icmmcce-15.2015.538
ID  - Meng2015/12
ER  -