Automatic Unpaired Shape Deformation Transfer

doi:10.1145/3272127.3275028

CORC > 计算技术研究所 > 中国科学院计算技术研究所 > 中国科学院计算技术研究所期刊论文 > 英文

	Automatic Unpaired Shape Deformation Transfer
	Qiao, Yi-Ling 3,4; Xia, Shihong 5; Xu, Weiwei 1; Gao, Lin 5; Rosin, Paul L.2; Lai, Yu-Kun 2; Yang, Jie 4,5
刊名	ACM TRANSACTIONS ON GRAPHICS
	2018-11-01
卷号	37 期号:6 页码:15
关键词	Deformation transfer generative adversarial network cycle consistency visual similarity
ISSN号	0730-0301
DOI	10.1145/3272127.3275028
英文摘要	Transferring deformation from a source shape to a target shape is a very useful technique in computer graphics. State-of-the-art deformation transfer methods require either point-wise correspondences between source and target shapes, or pairs of deformed source and target shapes with corresponding deformations. However, in most cases, such correspondences are not available and cannot be reliably established using an automatic algorithm. Therefore, substantial user effort is needed to label the correspondences or to obtain and specify such shape sets. In this work, we propose a novel approach to automatic deformation transfer between two unpaired shape sets without correspondences. 3D deformation is represented in a high-dimensional space. To obtain a more compact and effective representation, two convolutional variational autoencoders are learned to encode source and target shapes to their latent spaces. We exploit a Generative Adversarial Network (GAN) to map deformed source shapes to deformed target shapes, both in the latent spaces, which ensures the obtained shapes from the mapping are indistinguishable from the target shapes. This is still an under-constrained problem, so we further utilize a reverse mapping from target shapes to source shapes and incorporate cycle consistency loss, i.e. applying both mappings should reverse to the input shape. This VAE-Cycle GAN (VC-GAN) architecture is used to build a reliable mapping between shape spaces. Finally, a similarity constraint is employed to ensure the mapping is consistent with visual similarity, achieved by learning a similarity neural network that takes the embedding vectors from the source and target latent spaces and predicts the light field distance between the corresponding shapes. Experimental results show that our fully automatic method is able to obtain high-quality deformation transfer results with unpaired data sets, comparable or better than existing methods where strict correspondences are required.
资助项目	National Natural Science Foundation of China[61872440] ; National Natural Science Foundation of China[61828204] ; National Natural Science Foundation of China[61502453] ; National Natural Science Foundation of China[61772499] ; National Natural Science Foundation of China[61732016] ; Young Elite Scientists Sponsorship Program, CAST[2017QNRC001] ; Royal Society-Newton Mobility Grant[IE150731] ; CCF-Tencent Open Fund ; NVIDIA Corporation ; Alibaba IDEA Lab ; fundamental research fund for the central universities[2017YFB1002600]
WOS研究方向	Computer Science
语种	英语
出版者	ASSOC COMPUTING MACHINERY
WOS记录号	WOS:000455953100060
内容类型	期刊论文
源URL	[http://119.78.100.204/handle/2XEOYT63/3480]
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Xia, Shihong
作者单位	1.Zhejiang Univ, State Key Lab CAD&CG, Hangzhou, Zhejiang, Peoples R China 2.Cardiff Univ, Sch Comp Sci & Informat, Cardiff, S Glam, Wales 3.Chinese Acad Sci, Inst Comp Technol, Beijing, Peoples R China 4.Univ Chinese Acad Sci, Beijing, Peoples R China 5.Chinese Acad Sci, Inst Comp Technol, Beijing Key Lab Mobile Comp & Pervas Device, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Qiao, Yi-Ling,Xia, Shihong,Xu, Weiwei,et al. Automatic Unpaired Shape Deformation Transfer[J]. ACM TRANSACTIONS ON GRAPHICS,2018,37(6):15.
APA	Qiao, Yi-Ling.,Xia, Shihong.,Xu, Weiwei.,Gao, Lin.,Rosin, Paul L..,...&Yang, Jie.(2018).Automatic Unpaired Shape Deformation Transfer.ACM TRANSACTIONS ON GRAPHICS,37(6),15.
MLA	Qiao, Yi-Ling,et al."Automatic Unpaired Shape Deformation Transfer".ACM TRANSACTIONS ON GRAPHICS 37.6(2018):15.