CAGNet: Content-Aware Guidance for Salient Object Detection

作者：

Highlights：

• A Content-Aware Guidance Network for Salient Object Detection is introduced.

• The diverse recognition abilities of multi-level features are exploited to guide the features.

• Powerful multi-scale features are extracted by enabling densely connections within large regions in the feature maps.

• Our designed loss function outperforms the widely-used Cross-entropy loss by a large margin.

• Our method achieves the state-of-the-art performance on five challenging datasets.

摘要

•A Content-Aware Guidance Network for Salient Object Detection is introduced.•The diverse recognition abilities of multi-level features are exploited to guide the features.•Powerful multi-scale features are extracted by enabling densely connections within large regions in the feature maps.•Our designed loss function outperforms the widely-used Cross-entropy loss by a large margin.•Our method achieves the state-of-the-art performance on five challenging datasets.

论文关键词：Saliency detection,Fully convolutional neural networks,Attention guidance

论文评审过程：Received 9 October 2019, Revised 11 February 2020, Accepted 23 February 2020, Available online 24 February 2020, Version of Record 1 April 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107303