Attention Region Latent SVM for Image Classification

DOI

10.2991/isci-15.2015.328How to use a DOI?

Keywords

saliency map; SVM; image classification; optimization problem.

Abstract

This paper presents a new method for image classification based on image saliency region. The proposed attention region latent SVM (ARLSVM) is highly distinctive by training in a weakly-supervised manner which without requiring objects position or bounding boxes in training images. We use a latent SVM to model the optimization problem with saliency regions are latent variables. An EM method is proposed to solve the semi-convex optimization problem. Through experiments, our proposed approach performs favourably compared with two well-known algorithms in a benchmark dataset.

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  - Shengan Zhou
AU  - Peng Liang
AU  - Jiangwei Qin
PY  - 2015/01
DA  - 2015/01
TI  - Attention Region Latent SVM for Image Classification
BT  - Proceedings of the 2015 International Symposium on Computers & Informatics
PB  - Atlantis Press
SP  - 2532
EP  - 2539
SN  - 2352-538X
UR  - https://doi.org/10.2991/isci-15.2015.328
DO  - 10.2991/isci-15.2015.328
ID  - Zhou2015/01
ER  -