Spatial structure preserving feature pyramid network for semantic image segmentation
Yuan, Yuan1; Fang, Jie2,3; Lu, Xiaoqiang3; Feng, Yachuang3
刊名ACM Transactions on Multimedia Computing, Communications and Applications
2019-08
卷号15期号:3
ISSN号15516857;15516865
DOI10.1145/3321512
产权排序3
英文摘要

Recently, progress on semantic image segmentation is substantial, benefiting from the rapid development of Convolutional Neural Networks. Semantic image segmentation approaches proposed lately have been mostly based on Fully convolutional Networks (FCNs). However, these FCN-based methods use large receptive fields and too many pooling layers to depict the discriminative semantic information of the images. Specifically, on one hand, convolutional kernel with large receptive field smooth the detailed edges, since too much contexture information is used to depict the "center pixel." However, the pooling layer increases the receptive field through zooming out the latest feature maps, which loses many detailed information of the image, especially in the deeper layers of the network. These operations often cause low spatial resolution inside deep layers, which leads to spatially fragmented prediction. To address this problem, we exploit the inherent multi-scale and pyramidal hierarchy of deep convolutional networks to extract the feature maps with different resolutions and take full advantages of these feature maps via a gradually stacked fusing way. Specifically, for two adjacent convolutional layers, we upsample the features from deeper layer with stride of 2 and then stack them on the features from shallower layer. Then, a convolutional layer with kernels of 1 × 1 is followed to fuse these stacked features. The fused feature preserves the spatial structure information of the image; meanwhile, it owns strong discriminative capability for pixel classification. Additionally, to further preserve the spatial structure information and regional connectivity of the predicted category label map, we propose a novel loss term for the network. In detail, two graph model-based spatial affinity matrixes are proposed, which are used to depict the pixel-level relationships in the input image and predicted category label map respectively, and then their cosine distance is backward propagated to the network. The proposed architecture, called spatial structure preserving feature pyramid network, significantly improves the spatial resolution of the predicted category label map for semantic image segmentation. The proposed method achieves state-of-the-art results on three public and challenging datasets for semantic image segmentation. © 2019 Association for Computing Machinery.

语种英语
出版者Association for Computing Machinery
内容类型期刊论文
源URL[http://ir.opt.ac.cn/handle/181661/31848]  
专题西安光学精密机械研究所_光学影像学习与分析中心
通讯作者Lu, Xiaoqiang
作者单位1.Center for OPTical Imagery Analysis and Learning (OPTIMAL), Northwestern Polytechnical University, Xi'an, Shaanxi; 710072, China;
2.University of Chinese Academy of Sciences, Beijing; 100049, China;
3.Key Laboratory of Spectral Imaging Technology CAS, Xi'an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi'an, Shaanxi; 710119, China
推荐引用方式
GB/T 7714
Yuan, Yuan,Fang, Jie,Lu, Xiaoqiang,et al. Spatial structure preserving feature pyramid network for semantic image segmentation[J]. ACM Transactions on Multimedia Computing, Communications and Applications,2019,15(3).
APA Yuan, Yuan,Fang, Jie,Lu, Xiaoqiang,&Feng, Yachuang.(2019).Spatial structure preserving feature pyramid network for semantic image segmentation.ACM Transactions on Multimedia Computing, Communications and Applications,15(3).
MLA Yuan, Yuan,et al."Spatial structure preserving feature pyramid network for semantic image segmentation".ACM Transactions on Multimedia Computing, Communications and Applications 15.3(2019).
个性服务
查看访问统计
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。


©版权所有 ©2017 CSpace - Powered by CSpace