Abstract
Panoptic segmentation incorporating semantic segmentation and instance segmentation plays an important role in scene understanding. Although current research has made remarkable progress, it has great potential for improvement in the contour region. In this paper, we propose a Contour-Aware Panoptic Segmentation Network (CAPSNet) with a panoptic contour branch and a new structure loss. The panoptic contour branch is designed to enhance structural cues perception by guiding feature extraction. The novel structure loss is presented for associating the consistency between panoptic segmentation and contour. Experimental results on the Cityscapes dataset show that CAPSNet achieves 60.0 in PQ metric, improving by 0.9% over baseline, and the proposed panoptic contour branch and structure loss are able to realize promising boosts on panoptic segmentation task.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
