Salient object detection via double random walks with dual restarts

作者：

Highlights：

• We introduce the double-random-walks framework to detect salient objects.

• We define dual starts to consider the interacting and weighting of two agents.

• We present propagation distance to overcome the problem of geodesic distance.

• The proposed algorithm performs favorably against the existing saliency methods.

摘要

•We introduce the double-random-walks framework to detect salient objects.•We define dual starts to consider the interacting and weighting of two agents.•We present propagation distance to overcome the problem of geodesic distance.•The proposed algorithm performs favorably against the existing saliency methods.

论文关键词：Salient object detection,Double random walks,Propagation distance,Dual restarts

论文评审过程：Received 13 April 2018, Revised 17 September 2019, Accepted 23 October 2019, Available online 6 November 2019, Version of Record 21 January 2020.

论文官网地址：https://doi.org/10.1016/j.imavis.2019.10.008