跳到主要导航 跳到搜索 跳到主要内容

Co-saliency detection via looking deep and wide

  • Northwestern Polytechnical University Xian
  • Microsoft USA

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

123 引用 (Scopus)

摘要

With the goal of effectively identifying common and salient objects in a group of relevant images, co-saliency detection has become essential for many applications such as video foreground extraction, surveillance, image retrieval, and image annotation. In this paper, we propose a unified co-saliency detection framework by introducing two novel insights: 1) looking deep to transfer higher-level representations by using the convolutional neural network with additional adaptive layers could better reflect the properties of the co-salient objects, especially their consistency among the image group; 2) looking wide to take advantage of the visually similar neighbors beyond a certain image group could effectively suppress the influence of the common background regions when formulating the intra-group consistency. In the proposed framework, the wide and deep information are explored for the object proposal windows extracted in each image, and the co-saliency scores are calculated by integrating the intra-image contrast and intra-group consistency via a principled Bayesian formulation. Finally the window-level co-saliency scores are converted to the superpixel-level co-saliency maps through a foreground region agreement strategy. Comprehensive experiments on two benchmark datasets have demonstrated the consistent performance gain of the proposed approach.

源语言英语
主期刊名IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015
出版商IEEE Computer Society
页2994-3002
页数9
ISBN(电子版)9781467369640
DOI
出版状态已出版 - 14 10月 2015
活动IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015 - Boston, 美国
期限: 7 6月 2015 → 12 6月 2015

丛书

姓名Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
卷07-12-June-2015
ISSN(印刷版)1063-6919

会议

会议IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015
国家/地区美国
市Boston
时期7/06/15 → 12/06/15

学术指纹

探究 'Co-saliency detection via looking deep and wide' 的科研主题。它们共同构成独一无二的学术指纹。

引用此