Aggregate interactive learning for RGB-D salient object detection

作者：

Highlights：

• Proposes a strategy of aggregation and interaction to extract features.

• The deformable convolutional network is used to solve multi-scale problems.

• The one to one feature fusion module is used to realize the fusion of features.

摘要

•Proposes a strategy of aggregation and interaction to extract features.•The deformable convolutional network is used to solve multi-scale problems.•The one to one feature fusion module is used to realize the fusion of features.

论文关键词：Neural network,Salient object detection,Deformable convolution,Feature fusion module

论文评审过程：Received 26 October 2021, Revised 5 January 2022, Accepted 22 January 2022, Available online 10 February 2022, Version of Record 14 February 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116614