Quality-aware dual-modal saliency detection via deep reinforcement learning

作者：

Highlights：

• We introduce a novel and general quality estimation network using DRL.

• We apply it on the dual-modal saliency detection successfully and further propose a coarse-to-fine salient object detection framework based on GAN.

• Extensive experiments on two public dual-modal saliency detection dataset validated the effectiveness of the introduced algorithm.

摘要

•We introduce a novel and general quality estimation network using DRL.•We apply it on the dual-modal saliency detection successfully and further propose a coarse-to-fine salient object detection framework based on GAN.•Extensive experiments on two public dual-modal saliency detection dataset validated the effectiveness of the introduced algorithm.

论文关键词：Dual-modal saliency detection,Deep reinforcement learning,Quality-aware fusion,Generative adversarial networks

论文评审过程：Received 11 September 2018, Revised 16 March 2019, Accepted 16 March 2019, Available online 22 March 2019, Version of Record 16 April 2019.

论文官网地址：https://doi.org/10.1016/j.image.2019.03.012