Abstract
Fully convolutional networks (FCNs) have been widely applied for dense classification tasks such as semantic segmentation. As a large number of works based on FCNs are proposed, various semantic segmentation models have been improved significantly. However, duplicated upsampling and deconvolution operations in the FCNs will lead to information loss in semantic segmentation tasks and to problems such as ignoring the relationship between pixels and pixels and the lack of spatial consistency. In this study, we propose a parallel fully convolutional neural network that integrates holistically-nested edge detection (HED) network to capture image edge information, improving semantic segmentation performance. We carry out comprehensive experiments and achieve a better result on the PASCAL VOC 2012, PASCAL-Context and Cityscapes, comparing the results with some existing semantic segmentation methods.
Highlights
Semantic segmentation is a fundamental problem in image understanding [1] and in video interpretation [2]–[4]
To address the challenges of low segmentation accuracy in Fully convolutional networks (FCNs) and high computational cost in DeepLab, we introduce a simple, yet efficient parallel fully convolutional network (PFCN) for image semantic segmentation
We propose a parallel fully convolutional neural network consisting of two network branches
Summary
Semantic segmentation is a fundamental problem in image understanding [1] and in video interpretation [2]–[4]. Chen et al [10] came up with DeepLab that adopted atrous convolutions to expand receptive fields without reducing the resolution of the feature map and applied the Dense-CRF post-processing operations to refine the coarse FCN semantic segmentation This method has been widely applied in semantic segmentation task and achieved state-of-the-art performance. Some popular approaches generate high-resolution predictions by taking advantage of the features maps produced by shallow and middle layers, such as the FCN method in [9], DenseASPP in [11], and multi-label segmentation in [13] The purpose of these works is to obtain more detailed information from.
Sign-in/Register to view highlights & summary 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.
