基于深度卷积神经网络的弱监督图像语义分割

被引：6

作者：

郑宝玉 ^{[1
,2
]}

王雨 ^{[2
]}

吴锦雯 ^{[3
]}

周全 ^{[2
]}

机构：

[1] 南京邮电大学通信与信息工程学院

[2] 南京邮电大学通信与网络工程研究中心

[3] 中国地质大学(武汉)信息工程学院

来源：

南京邮电大学学报(自然科学版) | 2018年 / 38卷 / 05期

关键词：

语义分割; 深度卷积神经网络; 弱监督语义分割; 图像标注;

D O I：

10.14132/j.cnki.1673-5439.2018.05.001

中图分类号：

TP391.41 []; TP183 [人工神经网络与计算];

学科分类号：

080203 ;

摘要：

图像语义分割是计算机视觉领域重要识别任务,其目标是估计图像中的像素级目标类标签。最近,深度卷积神经网络(Deep Convolutional Neural Networks,DCNNs)已经成为解决图像语义分割的主流方法。然而,学习DCNNs需要大量的已标注训练数据(Ground Truth,GT),而现有数据集中的GT在数量和多样性方面因标注成本巨大而受到诸多限制。弱监督方法则考虑利用图像级标签和物体框之类的弱标注信息解决图像语义分割中的标注问题。相比于全监督的像素级图像标注,图像分类的GT(图像级标签)和目标检测的GT(物体框)更容易获得,因而可以直接借用为弱标注信息训练分类模型。弱监督语义分割的主要挑战在于标注信息的不完整性,即缺失了物体精确的边界信息。文中对基于DCNNs的弱监督语义分割方法进行了全面的阐述,描述了如何克服这些限制并讨论了提高其性能的可能研究方向。

引用

页码：1 / 12

页数：12

共 77 条

[41]

STC:A simple to complex framework for weaklysupervised semantic segmentation. Wei Y,Liang X,Chen Y,Shen X,Cheng MM,Feng JS,Zhao Y,Yan SC. IEEE Transactions on Pattern Analysis and Machine Intelligence . 2017

[42]

Boxsup:Exploiting bounding boxes to supervise convolutional networks for semantic segmentation. J.F.Dai,K.M.He,J.Sun. Proceedings of IEEE International Conference on Computer Vision . 2015

[43]

Built-in foreground/background prior for weakly-supervised semantic segmentation. Saleh F,Akbarian M S A,Salzmann M,et al. Proceedings of the 14th European Conference on Computer Vision . 2016

[44]

"Scribble Sup:scribble-supervised convolutional networks for semantic segmentation,". D.Lin,J.Dai,J.Jia,K.He,J.Sun. IEEE Conference on Computer Vision and Pattern Recognition . 2016

[45]

Learning transferrable knowledge for semantic segmentation with deep convolutional neural network. Hong,S,Oh,J,Lee,H,Han B. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition . 2016

[46]

Scene parsing through ADE20K dataset. Zhou B,Zhao H,Puig X,et al. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition . 2017

[47]

The Cityscapes Dataset for Semantic Urban Scene Understanding. Cordts M,Omran M,Ramos S,et al. Computer Vision and Pattern Recognition . 2016

[48]

Semantic contours from inverse detectors. Hariharan B,Arbelaez P,Bourdev L, et al. IEEE International Conference on Computer Vision . 2011

[49]

Learning object class detectors from weakly annotated video. Prest A,Leistner C,Civera J,et al. Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition . 2012

[50]

Augmented Feedback in Semantic Segmentation Under Image Level Supervision. X.J.Qi,et al. Computer Vision-ECCV 2016 . 2016

← 1 2 3 4 5 6 7 8 →