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

Zhou, Sanping; Wang, Jinjun; Wang, Le; Zhang, Jimuyang; Wang, Fei; Huang, Dong; Zheng, Nanning

doi:10.1109/tip.2020.3027992

Cited by 34 publications

(8 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, we design a bidirectional residual refinement module (BRRM) that captures the interaction between multiple layers to enhance the structural integrity and boundary retention of initial saliency predictions. Different from existing refinement methods that integrate multilevel information in intermediate salient object predictions, 26 – 29 we use two sequential residual blocks to refine the error of initial saliency predictions by way of bidirectional refinement. To capture the interaction of different layers, BRRM designs two additional processes, i.e., a forward–backward interaction process and a backward–forward interaction process.…”

Section: Methodsmentioning

confidence: 99%

“…In addition, existing attempts have introduced various techniques for achieving saliency prediction refinement 25 – 29 However, these methods either apply additional encoder–decoder refinement modules or use iterative decoder refinement blocks to extract multilevel information along with the low- and high-level features, respectively. This operation, which divides the backbone network into high- and low-level features, may make it difficult to exploit the information between adjacent levels.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Attention-based pyramid decoder network for salient object detection in remote sensing images

Liu

Lin

Yue

et al. 2022

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

Salient object detection (SOD) in remote sensing images (RSIs) is a highly practical task. However, scale variations of salient objects and the diversity of salient objects in RSIs pose challenges for detection. To address these issues, an attention-based pyramid decoder network (APDNet) is proposed for SOD in RSIs. The APDNet consists of three key components. First, a multiscale attention block is constructed to extract multiscale information and relations between salient objects, suppressing the distraction of variations of object types and scales. Second, a pyramid decoder structure is designed to take full advantage of multilevel features by gradually fusing features from two adjacent layers for result prediction. This feature fusion allows APDNet to efficiently exploit multilevel feature information, thus enabling a better feature representation. Finally, a bidirectional residual refinement module is proposed to enhance the structural integrity and boundary retention of initial saliency predictions. Extensive experiments on two public datasets demonstrate the superiority and effectiveness of the proposed APDNet against other compared state-of-the-art methods.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Attention-based pyramid decoder network for salient object detection in remote sensing images

Liu

Lin

Yue

et al. 2022

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

show abstract

“…Zhou et al. [43] designed a multi‐stage and dual‐path network structure to estimate the salient edges and regions from the low‐level and high‐level feature maps, respectively. Although these methods raise the bar of SOD greatly in terms of boundary recovery accuracy, there is still a large room for improving the fineness of salient maps.…”

Section: Related Workmentioning

confidence: 99%

“…Li et al [42] proposed an edge information-guided hierarchical feature fusion network, which fused features hierarchically and retained accurate semantic information and clear edge information effectively. Zhou et al [43] designed a multi-stage and dual-path network structure to estimate the salient edges and regions from the low-level and high-level feature maps, respectively. Although these methods raise the bar of SOD greatly in terms of boundary recovery accuracy, there is still a large room for improving the fineness of salient maps.…”

Section: Edge-aware Modelsmentioning

confidence: 99%

Salient object detection based on edge‐interior feature fusion

Shi

Qin

Liang

et al. 2022

IET Image Processing

View full text Add to dashboard Cite

Recently, existing FCNs‐based methods have shown their advantages in processing object boundaries. However, these methods still suffer from false object interference, which appears in saliency predictions. To solve this problem, an edge‐interior feature fusion (EIFF) framework is proposed, which consists of an internal‐boundary decoupled generation structure with receptive field enlargement and attention mechanism enhancement, and a salient feature refinement module. Specifically, the framework first learns edge features and interior features through an internal‐boundary decoupling generation network, which is supervised by labels obtained by decoupling ground‐truth through an image erosion algorithm. Then, feature refinement module (FRM) is designed to purify the coarse prediction by focusing on the ambiguous regions through a mining strategy to generate the final saliency map. To compensate for shortcomings of the BCE and IU loss, we also introduce a weighted loss to guide our model to focus more on the error‐prone parts. Experimental results on five benchmark datasets demonstrate that the proposed method performs favorably against 19 state‐of‐the‐art approaches under four standard metrics.

show abstract

“…Li et al [11] propose an edge information-guided hierarchical feature fusion network. Zhou et al [32] designed a multi-stage and dual-path network to jointly learn the salient edges and regions, in which the region branch network and edge branch network can interactively learn from each other. These methods improve the boundary prediction of the salient object to some extent, but the overall positioning lacks novel improvement.…”

Section: Related Workmentioning

confidence: 99%

Multi-scale Edge-based U-shape Network for Salient Object Detection

Sun¹,

Bian²,

Liu³

et al. 2021

Preprint

View full text Add to dashboard Cite

Deep-learning based salient object detection methods achieve great improvements. However, there are still problems existing in the predictions, such as blurry boundary and inaccurate location, which is mainly caused by inadequate feature extraction and integration. In this paper, we propose a Multi-scale Edge-based U-shape Network (MEUN) to integrate various features at different scales to achieve better performance. To extract more useful information for boundary prediction, U-shape Edge Network modules are embedded in each decoder units. Besides, the additional down-sampling module alleviates the location inaccuracy. Experimental results on four benchmark datasets demonstrate the validity and reliability of the proposed method. Multi-scale Edgebased U-shape Network also shows its superiority when compared with 15 state-of-the-art salient object detection methods.

show abstract

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

Cited by 34 publications

References 60 publications

Attention-based pyramid decoder network for salient object detection in remote sensing images

Attention-based pyramid decoder network for salient object detection in remote sensing images

Salient object detection based on edge‐interior feature fusion

Multi-scale Edge-based U-shape Network for Salient Object Detection

Contact Info

Product

Resources

About