.png){kind=icon}
返回
Deep layer guided network for salient object detection
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2019.09.018.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Recently salient object detection with convolutional neural networks has made great progress. More and more methods design more complex networks to integrate the features of each stage from backbone extractor. Considering that global information and spatial details of an image are better captured by features of deep layers and shallow layers of CNN respectively, deep layer guided network in which global information from deep layers is progressively transmitted to shallow layers in a guided manner is proposed. Hybrid feature enhancement block receives the feature maps of adjacent stages, and outputs enhanced feature maps which reduce the loss of spatial details and reduce the impact of varying in shape, scale and position of object. Discriminative feature block highlights the consistency in different stages from channel and spatial dimensions. Optimized discriminative features focus more on channels which show higher response to salient objects and positions consistent with the foreground in the saliency map from higher stage. Saliency inference block optimizes feature by merging with adjacent layer feature further. Thus feature becomes more enhanced, discriminative and refined in a high-to-low manner. Experimental results on five public datasets show that our salient object detection method reaches state-of-the-art performance under evaluation metrics. (C) 2019 Elsevier B.V. All rights reserved.
Keyword:
Salient object detection
Fully convolution network
Deep layer guidance
Discriminative feature
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
6.5
论文数:
2.5W
被引数:
6.5W
机构
引用论文
分享收藏分享收藏Deep adaptive feature embedding with local sample distributions for person re-identification.png)
PATTERN RECOGNITION
OAAI
PATTERN RECOGNITIONIF7.6
OAAI分享收藏Salient object detection for RGB-D image by single stream recurrent convolution neural network基于单流递归卷积神经网络的rgb-d图像显著性目标检测NEUROCOMPUTING
NEUROCOMPUTINGIF6.5
PREAI
分享收藏
查看更多内容