Hierarchical and Interactive Refinement Network for Edge-Preserving Salient Object Detection

被引:28
作者
Zhou, Sanping [1 ]
Wang, Jinjun [1 ]
Wang, Le [1 ]
Zhang, Jimuyang [2 ]
Wang, Fei [3 ]
Huang, Dong [2 ]
Zheng, Nanning [1 ]
机构
[1] Xi An Jiao Tong Univ, Inst Artificial Intelligence & Robot, Xian 710049, Peoples R China
[2] Carnegie Mellon Univ, Inst Robot, Pittsburgh, PA 15213 USA
[3] Xi An Jiao Tong Univ, Sch Software Engn, Xian 710049, Peoples R China
基金
中国国家自然科学基金;
关键词
Image edge detection; Object detection; Feature extraction; Inference algorithms; Training; Prediction algorithms; Semantics; Salient object detection; edge-guided inference; Hierarchical and Interactive Refinement Network; NEURAL-NETWORK;
D O I
10.1109/TIP.2020.3027992
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Salient object detection has undergone a very rapid development with the blooming of Deep Neural Network (DNN), which is usually taken as an important preprocessing procedure in various computer vision tasks. However, the down-sampling operations, such as pooling and striding, always make the final predictions blurred at edges, which has seriously degenerated the performance of salient object detection. In this paper, we propose a simple yet effective approach, i.e., Hierarchical and Interactive Refinement Network (HIRN), to preserve the edge structures in detecting salient objects. In particular, a novel multi-stage and dual-path network structure is designed to estimate the salient edges and regions from the low-level and high-level feature maps, respectively. As a result, the predicted regions will become more accurate by enhancing the weak responses at edges, while the predicted edges will become more semantic by suppressing the false positives in background. Once the salient maps of edges and regions are obtained at the output layers, a novel edge-guided inference algorithm is introduced to further filter the resulting regions along the predicted edges. Extensive experiments on several benchmark datasets have been conducted, in which the results show that our method significantly outperforms a variety of state-of-the-art approaches.
引用
收藏
页码:1 / 14
页数:14
相关论文
共 72 条
  • [21] DHSNet: Deep Hierarchical Saliency Network for Salient Object Detection
    Liu, Nian
    Han, Junwei
    [J]. 2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2016, : 678 - 686
  • [22] Liu Y, 2018, ARXIV180402864
  • [23] Employing Deep Part-Object Relationships for Salient Object Detection
    Liu, Yi
    Zhang, Qiang
    Zhang, Dingwen
    Han, Jungong
    [J]. 2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2019), 2019, : 1232 - 1241
  • [24] Liu Y, 2016, PROC CVPR IEEE, P231, DOI 10.1109/CVPR.2016.32
  • [25] Liu Yun, 2019, arXiv preprint arXiv:1903.12476
  • [26] How to Evaluate Foreground Maps?
    Margolin, Ran
    Zelnik-Manor, Lihi
    Tal, Ayellet
    [J]. 2014 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2014, : 248 - 255
  • [27] Mi JX, 2017, 2017 INTERNATIONAL CONFERENCE ON SECURITY, PATTERN ANALYSIS, AND CYBERNETICS (SPAC), P660, DOI 10.1109/SPAC.2017.8304358
  • [28] Movahedi V., 2010, P IEEE C COMP VIS PA, P49
  • [29] Learning to infer human attention in daily activities
    Nan, Zhixiong
    Shu, Tianmin
    Gong, Ran
    Wang, Shu
    Wei, Ping
    Zhu, Song-Chun
    Zheng, Nanning
    [J]. PATTERN RECOGNITION, 2020, 103
  • [30] Nan ZX, 2019, AAAI CONF ARTIF INTE, P8811