Learning Foreground-Background Segmentation from Improved Layered GANs

被引:6
作者
Yang, Yu [1 ]
Bilen, Hakan [2 ]
Zou, Qiran [1 ]
Cheung, Wing Yin [1 ]
Ji, Xiangyang [1 ]
机构
[1] Tsinghua Univ, BNRist, Beijing, Peoples R China
[2] Univ Edinburgh, Edinburgh, Midlothian, Scotland
来源
2022 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2022) | 2022年
基金
英国工程与自然科学研究理事会; 中国国家自然科学基金; 国家重点研发计划;
关键词
D O I
10.1109/WACV51458.2022.00044
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Deep learning approaches heavily rely on high-quality human supervision which is nonetheless expensive, time-consuming, and error-prone, especially for image segmentation task. In this paper, we propose a method to automatically synthesize paired photo-realistic images and segmentation masks for the use of training a foreground-background segmentation network. In particular, we learn a generative adversarial network that decomposes an image into foreground and background layers, and avoid trivial decompositions by maximizing mutual information between generated images and latent variables. The improved layered GANs can synthesize higher quality datasets from which segmentation networks of higher performance can be learned. Moreover, the segmentation networks are employed to stabilize the training of layered GANs in return, which are further alternately trained with Layered GANs. Experiments on a variety of single-object datasets show that our method achieves competitive generation quality and segmentation performance compared to related methods.
引用
收藏
页码:366 / 375
页数:10
相关论文
共 52 条
[1]  
Alexe B, 2010, LECT NOTES COMPUT SC, V6315, P380, DOI 10.1007/978-3-642-15555-0_28
[2]  
Ali Jahanian, 2020, ICLR
[3]  
Anciukevicius Titas, 2020, ARXIV PREPRINT ARXIV
[4]  
[Anonymous], 2011, PROC CVPR WORKSHOP F
[5]  
Arandjelovic R., 2019, ARXIV190511369
[6]  
Benny Yaniv, 2019, ECCV
[7]  
Bielski Adam, 2019, ADV NEUR IN, V32
[8]   FlyMap: Interacting with Maps Projected from a Drone [J].
Brock, Anke M. ;
Chatain, Julia ;
Park, Michelle ;
Fang, Tommy ;
Hachet, Martin ;
Landay, James A. ;
Cauchard, Jessica R. .
PROCEEDINGS PERVASIVE DISPLAYS 2018: THE 7TH ACM INTERNATIONAL SYMPOSIUM ON PERVASIVE DISPLAYS, 2018,
[9]  
Burgess Christopher P, 2019, Monet: Unsupervised scene decomposition and representation
[10]  
Chai YN, 2011, IEEE I CONF COMP VIS, P2579, DOI 10.1109/ICCV.2011.6126546