Image Inpainting by End-to-End Cascaded Refinement With Mask Awareness

doi:10.1109/TIP.2021.3076310

CASIA OpenIR > 模式识别实验室

	Image Inpainting by End-to-End Cascaded Refinement With Mask Awareness
	Zhu, Manyu 1,2; He, Dongliang 1; Li, Xin 1; Li, Chao1 ; Li, Fu 1; Liu, Xiao 1,3; Ding, Errui 1; Zhang, Zhaoxiang4
发表期刊	IEEE TRANSACTIONS ON IMAGE PROCESSING
ISSN	1057-7149
	2021
卷号	30 页码:4855-4866
通讯作者	He, Dongliang(hedlcc@126.com)
摘要	Inpainting arbitrary missing regions is challenging because learning valid features for various masked regions is nontrivial. Though U-shaped encoder-decoder frameworks have been witnessed to be successful, most of them share a common drawback of mask unawareness in feature extraction because all convolution windows (or regions), including those with various shapes of missing pixels, are treated equally and filtered with fixed learned kernels. To this end, we propose our novel mask-aware inpainting solution. Firstly, a Mask-Aware Dynamic Filtering (MADF) module is designed to effectively learn multi-scale features for missing regions in the encoding phase. Specifically, filters for each convolution window are generated from features of the corresponding region of the mask. The second fold of mask awareness is achieved by adopting Point-wise Normalization (PN) in our decoding phase, considering that statistical natures of features at masked points differentiate from those of unmasked points. The proposed PN can tackle this issue by dynamically assigning point-wise scaling factor and bias. Lastly, our model is designed to be an end-to-end cascaded refinement one. Supervision information such as reconstruction loss, perceptual loss and total variation loss is incrementally leveraged to boost the inpainting results from coarse to fine. Effectiveness of the proposed framework is validated both quantitatively and qualitatively via extensive experiments on three public datasets including Places2, CelebA and Paris StreetView.
关键词	Convolution Decoding Kernel Feature extraction Shape Image reconstruction Task analysis Image inpainting mask awareness dynamic filtering cascaded refinement
DOI	10.1109/TIP.2021.3076310
收录类别	SCI
语种	英语
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:000648333200007
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计	被引频次：48[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/45201
专题	模式识别实验室
通讯作者	He, Dongliang
作者单位	1.Baidu Inc, Dept Comp Vis VIS Technol, Beijing 100085, Peoples R China 2.ByteDance Inc, Beijing 100089, Peoples R China 3.TAL Educ Grp, Beijing 100080, Peoples R China 4.Chinese Acad Sci CASIA, Inst Automat, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Zhu, Manyu,He, Dongliang,Li, Xin,et al. Image Inpainting by End-to-End Cascaded Refinement With Mask Awareness[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2021,30:4855-4866.
APA	Zhu, Manyu.,He, Dongliang.,Li, Xin.,Li, Chao.,Li, Fu.,...&Zhang, Zhaoxiang.(2021).Image Inpainting by End-to-End Cascaded Refinement With Mask Awareness.IEEE TRANSACTIONS ON IMAGE PROCESSING,30,4855-4866.
MLA	Zhu, Manyu,et al."Image Inpainting by End-to-End Cascaded Refinement With Mask Awareness".IEEE TRANSACTIONS ON IMAGE PROCESSING 30(2021):4855-4866.