Key-point-guided adaptive convolution and instance normalization for continuous transitive face reenactment of any person

doi:10.1002/cav.2256

CORC > 自动化研究所 > 中国科学院自动化研究所 > 模式识别国家重点实验室 > 三维可视计算

	Key-point-guided adaptive convolution and instance normalization for continuous transitive face reenactment of any person
	Xu, Shibiao 3; Hua, Miao 2; Zhang, Jiguang 1; Zhang, Zhaohui 1; Zhang, Xiaopeng 1
刊名	COMPUTER ANIMATION AND VIRTUAL WORLDS
	2024-05-01
卷号	35 期号:3 页码:15
关键词	face reenactment human-centered computing visualization visualization application domains
ISSN号	1546-4261
DOI	10.1002/cav.2256
通讯作者	Zhang, Jiguang(jiguang.zhang@ia.ac.cn)
英文摘要	Face reenactment technology is widely applied in various applications. However, the reconstruction effects of existing methods are often not quite realistic enough. Thus, this paper proposes a progressive face reenactment method. First, to make full use of the key information, we propose adaptive convolution and instance normalization to encode the key information into all learnable parameters in the network, including the weights of the convolution kernels and the means and variances in the normalization layer. Second, we present continuous transitive facial expression generation according to all the weights of the network generated by the key points, resulting in the continuous change of the image generated by the network. Third, in contrast to classical convolution, we apply the combination of depth- and point-wise convolutions, which can greatly reduce the number of weights and improve the efficiency of training. Finally, we extend the proposed face reenactment method to the face editing application. Comprehensive experiments demonstrate the effectiveness of the proposed method, which can generate a clearer and more realistic face from any person and is more generic and applicable than other methods. This work presents a continuous transitive face reenactment algorithm that uses face key points information to gradually reenact faces based on two stages GAN, which contains the key face points transformation module and the facial expression generation module. The process involves transforming key points from the source face and generating corresponding facial expressions on the target face. image
资助项目	Beijing Natural Science Foundation ; National Natural Science Foundation of China[62271074] ; National Natural Science Foundation of China[62171321] ; National Natural Science Foundation of China[62162044] ; National Natural Science Foundation of China[52175493] ; National Natural Science Foundation of China[32271983] ; Open Project Program of State Key Laboratory of Virtual Reality Technology and Systems, Beihang University[VRLAB2023B01] ; Wenzhou Business School 2024 Talent launch program[RC202401] ; [JQ23014]
WOS关键词	RECONSTRUCTION
WOS研究方向	Computer Science
语种	英语
出版者	WILEY
WOS记录号	WOS:001230174100001
资助机构	Beijing Natural Science Foundation ; National Natural Science Foundation of China ; Open Project Program of State Key Laboratory of Virtual Reality Technology and Systems, Beihang University ; Wenzhou Business School 2024 Talent launch program
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/58440]
专题	模式识别国家重点实验室_三维可视计算
通讯作者	Zhang, Jiguang
作者单位	1.Chinese Acad Sci, Inst Automat, Beijing, Peoples R China 2.Beijing Bytedance Technol Co Ltd, Beijing, Peoples R China 3.Beijing Univ Posts & Telecommun, Sch Artificial Intelligence, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Xu, Shibiao,Hua, Miao,Zhang, Jiguang,et al. Key-point-guided adaptive convolution and instance normalization for continuous transitive face reenactment of any person[J]. COMPUTER ANIMATION AND VIRTUAL WORLDS,2024,35(3):15.
APA	Xu, Shibiao,Hua, Miao,Zhang, Jiguang,Zhang, Zhaohui,&Zhang, Xiaopeng.(2024).Key-point-guided adaptive convolution and instance normalization for continuous transitive face reenactment of any person.COMPUTER ANIMATION AND VIRTUAL WORLDS,35(3),15.
MLA	Xu, Shibiao,et al."Key-point-guided adaptive convolution and instance normalization for continuous transitive face reenactment of any person".COMPUTER ANIMATION AND VIRTUAL WORLDS 35.3(2024):15.