261 Hits in 4.7 sec

Image Inpainting for Irregular Holes Using Partial Convolutions [article]

Guilin Liu, Fitsum A. Reda, Kevin J. Shih, Ting-Chun Wang, Andrew Tao, Bryan Catanzaro
2018 arXiv   pre-print
Existing deep learning based image inpainting methods use a standard convolutional network over the corrupted image, using convolutional filter responses conditioned on both valid pixels as well as the  ...  We propose the use of partial convolutions, where the convolution is masked and renormalized to be conditioned on only valid pixels.  ...  We would like to thank Jonah Alben, Rafael Valle Costa, Karan Sapra, Chao Yang, Raul Puri, Brandon Rowlett and other NVIDIA colleagues for valuable discussions, and Chris Hebert for technical support.  ... 
arXiv:1804.07723v2 fatcat:xpa332jvtfgenbnzr6bph75n5m

A Benchmark for Inpainting of Clothing Images with Irregular Holes [article]

Furkan Kınlı, Barış Özcan, Furkan Kıraç
2020 arXiv   pre-print
For that matter, we present an extensive benchmark of clothing image inpainting on well-known fashion datasets.  ...  Furthermore, we introduce the use of a dilated version of partial convolutions, which efficiently derive the mask update step, and empirically show that the proposed method reduces the required number  ...  -We present the idea of using dilated version of partial convolutions for image inpainting tasks, where the dilated input window for partial convolutions derives more efficient mask update step.  ... 
arXiv:2007.05080v3 fatcat:dhvfhsaerbcwthc2fhomkr3kne

Inpainting CMB maps using Partial Convolutional Neural Networks [article]

Gabriele Montefalcone, Maximilian H. Abitbol, Darsh Kodwani, R.D.P. Grumitt
2020 arXiv   pre-print
We present a novel application of partial convolutional neural networks (PCNN) that can inpaint masked images of the cosmic microwave background.  ...  This should be particularly beneficial to inpaint irregular masks for the CMB that come from astrophysical sources such as galactic foregrounds.  ...  research using partial convolutions.  ... 
arXiv:2011.01433v3 fatcat:wyvkhmfm4zfrbbcyu2d5equ5fm

Image Inpainting with Learnable Bidirectional Attention Maps [article]

Chaohao Xie, Shaohui Liu, Chao Li, Ming-Ming Cheng, Wangmeng Zuo, Xiao Liu, Shilei Wen, Errui Ding
2019 arXiv   pre-print
Most convolutional network (CNN)-based inpainting methods adopt standard convolution to indistinguishably treat valid pixels and holes, making them limited in handling irregular holes and more likely to  ...  In this paper, we present a learnable attention map module for learning feature renormalization and mask-updating in an end-to-end manner, which is effective in adapting to irregular holes and propagation  ...  For better handling irregular holes and suppressing color discrepancy and blurriness, partial convolution (PConv) [17] has been suggested.  ... 
arXiv:1909.00968v3 fatcat:goyly3uez5gnbl2w5qu36ojfzy

Deep Inception Generative Network for Cognitive Image Inpainting [article]

Qingguo Xiao, Guangyao Li, Qiaochuan Chen
2018 arXiv   pre-print
Recent advances in deep learning have shown exciting promise in filling large holes and lead to another orientation for image inpainting.  ...  Experiments for regular, irregular, and custom regions completion are all performed and free-style image inpainting is also presented.  ...  [18] proposed the concept of partial convolutions. It has shown outstanding performance for images inpainting with irregular holes.  ... 
arXiv:1812.01458v1 fatcat:3zcwa53l6bhzpfvctopfzf5onq

Free-Form Image Inpainting with Gated Convolution [article]

Jiahui Yu, Zhe Lin, Jimei Yang, Xiaohui Shen, Xin Lu, Thomas Huang
2019 arXiv   pre-print
We present a generative image inpainting system to complete images with free-form mask and guidance.  ...  The proposed gated convolution solves the issue of vanilla convolution that treats all input pixels as valid ones, generalizes partial convolution by providing a learnable dynamic feature selection mechanism  ...  Gated Convolution We first explain why vanilla convolutions used in [15, 49] are ill-fitted for the task of free-form image inpainting.  ... 
arXiv:1806.03589v2 fatcat:3kv4yivy3fgd3mzpf7diimepii

Contextual Residual Aggregation for Ultra High-Resolution Image Inpainting [article]

Zili Yi, Qiang Tang, Shekoofeh Azizi, Daesik Jang, Zhan Xu
2020 arXiv   pre-print
Our model can inpaint images as large as 8K with considerable hole sizes, which is intractable with previous learning-based approaches.  ...  Recently data-driven image inpainting methods have made inspiring progress, impacting fundamental image editing tasks such as object removal and damaged image repairing.  ...  Acknowledgment We want to acknowledge Peng Deng, Shao Hua Chen, Xinjiang Sun, Chunhua Tian and other colleagues in Huawei Technologies for their support in the project.  ... 
arXiv:2005.09704v1 fatcat:qu4ixcjoobg7zpv5ocgcyumwhm

3DBooSTeR: 3D Body Shape and Texture Recovery [article]

Alexandre Saint, Anis Kacem, Kseniya Cherenkova, Djamila Aouada
2020 arXiv   pre-print
The texture is subsequently obtained by projecting the partial texture onto the template mesh before inpainting the corresponding texture map with a novel approach.  ...  With the advent of virtual and augmented reality, there is a demand for creating realistic and high-fidelity digital 3D human representations.  ...  It is specifically designed to handle holes of irregular shape and to be robust to irrelevant background image information.  ... 
arXiv:2010.12670v1 fatcat:4lbx3ppbkfgpzczxbz6blhtst4

Semi-parametric Image Inpainting [article]

Karim Iskakov
2018 arXiv   pre-print
This paper introduces a semi-parametric approach to image inpainting for irregular holes. The nonparametric part consists of an external image database.  ...  During test time database is used to retrieve a supplementary image, similar to the input masked picture, and utilize it as auxiliary information for the deep neural network.  ...  [16] showed stunning results, archiving state-of-the-art in image inpainting for irregular masks. Authors use modification of common convolution layer -partial convolution.  ... 
arXiv:1807.02855v2 fatcat:ib7g3vevpbhp7j44gfsa4ilhxi

Large Hole Image Inpainting With Compress-Decompression Network [article]

Zhenghang Wu, Yidong Cui
2020 arXiv   pre-print
Image inpainting technology can patch images with missing pixels. Existing methods propose convolutional neural networks to repair corrupted images.  ...  The compression network takes responsibility for inpainting and generating a down-sample image.  ...  [4] proposed Partial Convolution to modify the typical convolutional network, and the new network can well support irregular holes. Song et al.  ... 
arXiv:2002.00199v1 fatcat:oy6yakgjxrfoljp25uvhieamuy

Image Inpainting with Learnable Edge-Attention Maps

Liujie Sun, Qinghan Zhang, Wenju Wang, Mingxi Zhang
2020 IEEE Access  
This paper proposes an end-to-end Learnable Edge-Attention Map (LEAM) method to assist image inpainting.  ...  For more information, see VOLUME 9, 2021  ...  The authors would like to thank Deepak Pathak, the author of Context Encoders: Feature Learning by Inpainting, for providing the Paris StreetView data.  ... 
doi:10.1109/access.2020.3047740 fatcat:bhoesnc4ifakjfuokfpw2oln3y

Rethinking Image Inpainting via a Mutual Encoder-Decoder with Feature Equalizations [article]

Hongyu Liu, Bin Jiang, Yibing Song, Wei Huang, Chao Yang
2020 arXiv   pre-print
Deep encoder-decoder based CNNs have advanced image inpainting methods for hole filling.  ...  We use the equalized feature to supplement decoder features for output image generation through skip connections.  ...  This work is partially supported by the National Natural Science Foundation of China under Grant No. 61702176.  ... 
arXiv:2007.06929v1 fatcat:cqcgdig2sbcdhhvymehfc34lra

R-MNet: A Perceptual Adversarial Network for Image Inpainting [article]

Jireh Jam and Connah Kendrick and Vincent Drouard and Kevin Walker and Gee-Sern Hsu and Moi Hoon Yap
2020 arXiv   pre-print
We address the problem by proposing a Wasserstein GAN combined with a new reverse mask operator, namely Reverse Masking Network (R-MNet), a perceptual adversarial network for image inpainting.  ...  The reverse mask operator transfers the reverse masked image to the end of the encoder-decoder network leaving only valid pixels to be inpainted.  ...  We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Quadro P6000 used for this research.  ... 
arXiv:2008.04621v3 fatcat:zu3yupgft5dzphjcafddxohmle

Progressive Image Inpainting with Full-Resolution Residual Network [article]

Zongyu Guo, Zhibo Chen, Tao Yu, Jiale Chen, Sen Liu
2019 arXiv   pre-print
Recently, learning-based algorithms for image inpainting achieve remarkable progress dealing with squared or irregular holes.  ...  In this paper, we propose full-resolution residual network (FRRN) to fill irregular holes, which is proved to be effective for progressive image inpainting.  ...  Figure 1 : 1 Figure 1: Progressive image inpainting for irregular holes.  ... 
arXiv:1907.10478v3 fatcat:wlhkftndhzh7bk5wgxixbob2by

Coherent Semantic Attention for Image Inpainting [article]

Hongyu Liu and Bin Jiang and Yi Xiao and Chao Yang
2019 arXiv   pre-print
The latest deep learning-based approaches have shown promising results for the challenging task of inpainting missing regions of an image.  ...  The experiments on CelebA, Places2, and Paris StreetView datasets have validated the effectiveness of our proposed methods in image inpainting tasks and can obtain images with a higher quality as compared  ...  For irregular holes, we compare with PC and GC using Paris StreetView [8] and CelebA [27] validation images.  ... 
arXiv:1905.12384v3 fatcat:hbhjexe7f5fdxpsnrqylzzdbuq
« Previous Showing results 1 — 15 out of 261 results