Edge-aware salient object detection network via context guidance

作者：

Highlights：

• An edge-aware saliency detection network using context guidance is proposed.

• Semantics guides selective learning of local details in multi-level features.

• Supervising low-level edge information preserves boundaries of salient object.

• Our method performs favorably against the recent representative methods.

摘要

•An edge-aware saliency detection network using context guidance is proposed.•Semantics guides selective learning of local details in multi-level features.•Supervising low-level edge information preserves boundaries of salient object.•Our method performs favorably against the recent representative methods.

论文关键词：Salient object detection,Saliency,Context guidance,Attention,Multi-scale feature

论文评审过程：Received 17 February 2021, Revised 26 March 2021, Accepted 4 April 2021, Available online 8 April 2021, Version of Record 15 April 2021.

论文官网地址：https://doi.org/10.1016/j.imavis.2021.104166