Unsupervised Dense Deformation Embedding Network for Template-Free Shape Correspondence

CORC > 沈阳自动化研究所 > 中国科学院沈阳自动化研究所 > 机器人学研究室

	Unsupervised Dense Deformation Embedding Network for Template-Free Shape Correspondence
	Chen RH(陈荣瀚)1,2,3; Cong Y(丛杨)2,3; Dong JH(董家华)
	2021
会议日期	October 10-17, 2021
会议地点	Montreal, Canada
页码	8341-8350
英文摘要	Shape correspondence from 3D deformation learning has attracted appealing academy interests recently. Nevertheless, current deep learning based methods require the supervision of dense annotations to learn per-point translations, which severely over-parameterize the deformation process. Moreover, they fail to capture local geometric details of original shape via global feature embedding. To address these challenges, we develop a new Unsupervised Dense Deformation Embedding Network (i.e., UD2E-Net), which learns to predict deformations between non-rigid shapes from dense local features. Since it is non-trivial to match deformation-variant local features for deformation prediction, we develop an Extrinsic-Intrinsic Autoencoder to first encode extrinsic geometric features from source into intrinsic coordinates in a shared canonical shape, with which the decoder then synthesizes corresponding target features. Moreover, a bounded maximum mean discrepancy loss is developed to mitigate the distribution divergence between the synthesized and original features. To learn natural deformation without dense supervision, we introduce a coarse parameterized deformation graph, for which a novel trace and propagation algorithm is proposed to improve both the quality and efficiency of the deformation. Our UD2E-Net outperforms state-of-the-art unsupervised methods by 24% on Faust Inter challenge and even supervised methods by 13% on Faust Intra challenge.
产权排序	1
会议录	2021 IEEE/CVF International Conference on Computer Vision (ICCV)
会议录出版者	IEEE
会议录出版地	New York
语种	英语
ISSN号	1550-5499
ISBN号	978-1-6654-2812-5
内容类型	会议论文
源URL	[http://ir.sia.cn/handle/173321/29963]
专题	沈阳自动化研究所_机器人学研究室
通讯作者	Cong Y(丛杨)
作者单位	1.University of Chinese Academy of Sciences 2.State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences 3.Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences
推荐引用方式 GB/T 7714	Chen RH,Cong Y,Dong JH. Unsupervised Dense Deformation Embedding Network for Template-Free Shape Correspondence[C]. 见:. Montreal, Canada. October 10-17, 2021.