Adversarial Learning Semantic Volume for 2D/3D Face Shape Regression in the Wild

doi:10.1109/TIP.2019.2911114

CORC > 自动化研究所 > 中国科学院自动化研究所 > 智能感知与计算研究中心

	Adversarial Learning Semantic Volume for 2D/3D Face Shape Regression in the Wild
	Zhang, Hongwen1,2 ; Li, Qi3 ; Sun, Zhenan1,2
刊名	IEEE TRANSACTIONS ON IMAGE PROCESSING
	2019-09-01
卷号	28 期号:9 页码:4526-4540
关键词	2D/3D facial landmark localization semantic volumetric representation joint voxel and coordinate regression auxiliary regression adversarial learning
ISSN号	1057-7149
DOI	10.1109/TIP.2019.2911114
通讯作者	Sun, Zhenan(znsun@nlpr.ia.ac.cn)
英文摘要	Regression-based methods have revolutionized 2D landmark localization with the exploitation of deep neural networks and massive annotated datasets in the wild. However, it remains challenging for 3D landmark localization due to the lack of annotated datasets and the ambiguous nature of landmarks under the 3D perspective. This paper revisits regressionbased methods and proposes an adversarial voxel and coordinate regression framework for 2D and 3D facial landmark localization in real-world scenarios. First, a semantic volumetric representation is introduced to encode the per-voxel likelihood of positions being the 3D landmarks. Then, an end-to-end pipeline is designed to jointly regress the proposed volumetric representation and the coordinate vector. Such a pipeline not only enhances the robustness and accuracy of the predictions but also unifies the 2D and 3D landmark localization so that the 2D and 3D datasets could be utilized simultaneously. Further, an adversarial learning strategy is exploited to distill 3D structure learned from synthetic datasets to real-world datasets under weakly supervised settings, where an auxiliary regression discriminator is proposed to encourage the network to produce plausible predictions for both the synthetic and real-world images. The effectiveness of our method is validated on benchmark datasets 3DFAW and AFLW2000-3D for both 2D and 3D facial landmark localization tasks. The experimental results show that the proposed method achieves significant improvements over the previous state-of-the-art methods.
资助项目	National Natural Science Foundation of China[U1836217] ; National Natural Science Foundation of China[61427811] ; National Natural Science Foundation of China[61573360] ; National Natural Science Foundation of China[61721004] ; National Natural Science Foundation of China[61702513] ; National Natural Science Foundation of China[61806197] ; National Key Research and Development Program of China[2017YFC0821602] ; National Key Research and Development Program of China[2016YFB1001000]
WOS关键词	ALIGNMENT
WOS研究方向	Computer Science ; Engineering
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:000476797800006
资助机构	National Natural Science Foundation of China ; National Key Research and Development Program of China
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/26090]
专题	自动化研究所_智能感知与计算研究中心
通讯作者	Sun, Zhenan
作者单位	1.Chinese Acad Sci, CAS Ctr Excellence Brain Sci & Intelligence Techn, Ctr Res Intelligent Percept & Comp, Natl Lab Pattern Recognit,Inst Automat, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Beijing 101408, Peoples R China 3.Chinese Acad Sci, Ctr Res Intelligent Percept & Com, Natl Lab Pattern Recognit, Inst Automat, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Zhang, Hongwen,Li, Qi,Sun, Zhenan. Adversarial Learning Semantic Volume for 2D/3D Face Shape Regression in the Wild[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2019,28(9):4526-4540.
APA	Zhang, Hongwen,Li, Qi,&Sun, Zhenan.(2019).Adversarial Learning Semantic Volume for 2D/3D Face Shape Regression in the Wild.IEEE TRANSACTIONS ON IMAGE PROCESSING,28(9),4526-4540.
MLA	Zhang, Hongwen,et al."Adversarial Learning Semantic Volume for 2D/3D Face Shape Regression in the Wild".IEEE TRANSACTIONS ON IMAGE PROCESSING 28.9(2019):4526-4540.