Vol.26, No.4, 2020, pp.847-855, doi:10.32604/iasc.2020.010119
OPEN ACCESS
REVIEW
PGCA-Net: Progressively Aggregating Hierarchical Features with the Pyramid Guided Channel Attention for Saliency Detection
  • Jiajie Mai1, Xuemiao Xu2,*, Guorong Xiao3, Zijun Deng2, Jiaxing Chen2
1 Department of Telecommunications Engineering and Management, Beijing University of Posts and Telecommunication, Beijing100876, China
2 School of Computer Science and Engineering, South China University of Technology, Guangzhou 510006, China
3 Key Laboratory of Science & Technology and Finance, Guangdong University of Finance, Guangzhou 510521, China
* Corresponding Author: Xuemiao Xu, xuemx@scut.edu.cn
Abstract
The Salient object detection aims to segment out the most visually distinctive objects in an image, which is a challenging task in computer vision. In this paper, we present the PGCA-Net equipped with the pyramid guided channel attention fusion block (PGCAFB) for the saliency detection task. Given an input image, the hierarchical features are extracted using a deep convolutional neural network (DCNN), then starting from the highest-level semantic features, we stage-by-stage restore the spatial saliency details by aggregating the lowerlevel detailed features. Since for the weak discriminative ability of the shallow detailed features, directly introducing them to the semantic features will only lead to sub-optimal results. Thus, we take a novel pyramid channel attention mechanism to attend to the useful detailed shallow feature channels before aggregation. The experimental results show that our proposed method outperforms its competitors on 5 benchmark testing sets.
Keywords
Saliency detection; channel attention; image segmentation; computer vision; deep learning.
Cite This Article
J. Mai, X. Xu, G. Xiao, Z. Deng and J. Chen, "Pgca-net: progressively aggregating hierarchical features with the pyramid guided channel attention for saliency detection," Intelligent Automation & Soft Computing, vol. 26, no.4, pp. 847–855, 2020.
This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.