Saliency detection via background and foreground null space learning

作者：

Highlights：

• The importance of discriminative features with zero intra-class variance.

• The use of the null space learning.

• Measuring the distances in the null space as the rough estimation of the saliency.

• A two-stage cascade saliency detection scheme.

摘要

•The importance of discriminative features with zero intra-class variance.•The use of the null space learning.•Measuring the distances in the null space as the rough estimation of the saliency.•A two-stage cascade saliency detection scheme.

论文关键词：Saliency detection,Null space learning,Saliency optimization

论文评审过程：Received 5 March 2018, Revised 8 October 2018, Accepted 8 October 2018, Available online 29 October 2018, Version of Record 23 November 2018.

论文官网地址：https://doi.org/10.1016/j.image.2018.10.005