BiconNet: An edge-preserved connectivity-based approach for salient object detection. (January 2022)

• Record Type: Journal Article
• Authors: Yang, Ziyun; Soltanian-Zadeh, Somayyeh; Farsiu, Sina
• Is Part Of: Pattern recognition. Volume 121(2022)
• Publication Date: 2022-01
• Subjects: Salient object detection -- Visual saliency -- Connectivity modeling -- Deep learning -- Edge modeling; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2021.108231
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: A new connectivity-based CNN (BiconNet) for salient object detection is proposed. Inter-pixel relationship is modeled by using connectivity masks as training labels. Edge features are emphasized via a newly proposed and efficient module. BiconNet can be plugged into existing models with a neglectable parameter increase. BiconNet outperformed existing salient object detection methods. Abstract: Salient object detection (SOD) is viewed as a pixel-wise saliency modeling task by traditional deep learning-based methods. A limitation of current SOD models is insufficient utilization of inter-pixel information, which usually results in imperfect segmentation near edge regions and low spatial coherence. As we demonstrate, using a saliency mask as the only label is suboptimal. To address this limitation, we propose a connectivity-based approach called bilateral connectivity network (BiconNet), which uses connectivity masks together with saliency masks as labels for effective modeling of inter-pixel relationships and object saliency. Moreover, we propose a bilateral voting module to enhance the output connectivity map, and a novel edge feature enhancement method that efficiently utilizes edge-specific features. Through comprehensive experiments on five benchmark datasets, we demonstrate that our proposed method can be plugged into any existing state-of-the-art saliency-based SOD framework to improve its performance with negligible parameter increase.