International Journal of Computer Vision

, Volume 120, Issue 2, pp 215–232

Detection of Co-salient Objects by Looking Deep and Wide

  • Dingwen Zhang
  • Junwei Han
  • Chao Li
  • Jingdong Wang
  • Xuelong Li
Article

DOI: 10.1007/s11263-016-0907-4

Cite this article as:
Zhang, D., Han, J., Li, C. et al. Int J Comput Vis (2016) 120: 215. doi:10.1007/s11263-016-0907-4

Abstract

In this paper, we propose a unified co-salient object 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 sematic properties of the co-salient objects; (2) looking wide to take advantage of the visually similar neighbors from other image groups could effectively suppress the influence of the common background regions. The wide and deep information are explored for the object proposal windows extracted in each image. The window-level co-saliency scores are calculated by integrating the intra-image contrast, the intra-group consistency, and the inter-group separability via a principled Bayesian formulation and are then converted to the superpixel-level co-saliency maps through a foreground region agreement strategy. Comprehensive experiments on two existing and one newly established datasets have demonstrated the consistent performance gain of the proposed approach.

Keywords

Co-saliency detection Domain adaptive convolutional neural network Bayesian framework 

Copyright information

© Springer Science+Business Media New York 2016

Authors and Affiliations

  • Dingwen Zhang
    • 1
  • Junwei Han
    • 1
  • Chao Li
    • 1
  • Jingdong Wang
    • 2
  • Xuelong Li
    • 3
  1. 1.School of AutomationNorthwestern Polytechnical UniversityXi’anChina
  2. 2.Microsoft Research AsiaBeijingChina
  3. 3. Xi’an Institute of Optics and Precision MechanicsChinese Academy of SciencesXi’anChina