Share Email Print

Journal of Electronic Imaging

Image salient object detection with refined deep features via convolution neural network
Author(s): Dandan Zhu; Lei Dai; Xuan Shao; Qiangqiang Zhou; Laurent Itti; Ye Luo; Jianwei Lu
Format Member Price Non-Member Price
PDF $20.00 $25.00

Paper Abstract

Recent advances in saliency detection have used deep learning to obtain high-level features to detect salient regions. These advances have demonstrated superior results over previous works that use handcrafted low-level features for saliency detection. We propose a convolutional neural network (CNN) model to learn high-level features for saliency detection. Compared to other methods, our method presents two merits. First, when performing features extraction, apart from the convolution and pooling step in our method, we add restricted Boltzmann machine into the CNN framework to obtain more accurate features in intermediate step. Second, in order to avoid manual annotation data, we add deep belief network classifier at the end of this model to classify salient and nonsalient regions. Quantitative and qualitative experiments on three benchmark datasets demonstrate that our method performs favorably against the state-of-the-art methods.

Paper Details

Date Published: 23 November 2017
PDF: 11 pages
J. Electron. Imag. 26(6) 063018 doi: 10.1117/1.JEI.26.6.063018
Published in: Journal of Electronic Imaging Volume 26, Issue 6
Show Author Affiliations
Dandan Zhu, Tongji Univ. (China)
Lei Dai, Jiangsu Univ. (China)
Xuan Shao, Tongji Univ. (China)
Qiangqiang Zhou, Tongji Univ. (China)
Laurent Itti, The Univ. of Southern California (United States)
Ye Luo, Tongji Univ. (China)
Jianwei Lu, Tongji Univ. (China)
Advanced Institute of Translational Medicine, Tongji Univ. (China)

© SPIE. Terms of Use
Back to Top