Coarse-to-fine salient object detection based on deep convolutional neural networks

作者：

Highlights：

•

• We propose a coarse-to-fine approach for salient object detection.

• We use pixel-wise FCN to predict the coarse saliency map.

• A superpixel-based CNN is constructed to refine the detected salient objects with precise boundaries.

• Our method performs well in locating and segmenting the salient objects.

摘要

•We propose a coarse-to-fine approach for salient object detection.•We use pixel-wise FCN to predict the coarse saliency map.•A superpixel-based CNN is constructed to refine the detected salient objects with precise boundaries.•Our method performs well in locating and segmenting the salient objects.

论文关键词：Saliency detection,Convolutional neural network,Fully convolutional neural network

论文评审过程：Received 24 August 2017, Revised 27 January 2018, Accepted 31 January 2018, Available online 22 March 2018, Version of Record 22 March 2018.

论文官网地址：https://doi.org/10.1016/j.image.2018.01.012