Leveraging human fixations in sparse coding: Learning a discriminative dictionary for saliency prediction

Ming Jiang, Mingli Song, Qi Zhao

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Scopus citations

Abstract

This paper proposes to learn a discriminative dictionary for saliency detection. In addition to the conventional sparse coding mechanism that learns a representational dictionary of natural images for saliency prediction, this work uses supervised information from eye tracking experiments in training to enhance the discriminative power of the learned dictionary. Furthermore, we explicitly model saliency at multi-scale by formulating it as a multi-class problem, and a label consistency term is incorporated into the framework to encourage class (salient vs. non-salient) and scale consistency in the learned sparse codes. K-SVD is employed as the central computational module to efficiently obtain the optimal solution. Experiments demonstrate the superior performance of the proposed algorithm compared with the state-of-the-art in saliency prediction.

Original languageEnglish (US)
Title of host publicationProceedings - 2013 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2013
Pages2126-2133
Number of pages8
DOIs
StatePublished - Dec 1 2013
Event2013 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2013 - Manchester, United Kingdom
Duration: Oct 13 2013Oct 16 2013

Publication series

NameProceedings - 2013 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2013

Other

Other2013 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2013
CountryUnited Kingdom
CityManchester
Period10/13/1310/16/13

Keywords

  • Dictionary learning
  • K-svd
  • Saliency
  • Supervised sparse coding
  • Visual attention

Fingerprint Dive into the research topics of 'Leveraging human fixations in sparse coding: Learning a discriminative dictionary for saliency prediction'. Together they form a unique fingerprint.

Cite this