Filling-in by joint interpolation of vector fields and gray levels

By Coloma Ballester, Marcelo Bertalmío, V. Caselles, Guillermo Sapiro, Joan Verdera. 2001

1,065 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Computer Vision and Pattern Recognition

Citations in the corpus

Large Scale Image Completion via Co-Modulated Generative Adversarial Networks (2021, methodology): “Traditional methods (Ballester et al., 2001; Barnes et al., 2009; Darabi et al., 2012; Efros & Freeman, 2001; Efros & Leung, 1999) utilize only low-level features and fail to generate semantically consistent contents.”
Regionwise Generative Adversarial Image Inpainting for Large Missing Areas (2019, methodology): “The former methods propagate background data into missing regions by following a diffusive process typically modeled using differential operators [18], [19].”
EdgeConnect: Generative Image Inpainting with Adversarial Edge Learning (2019, methodology): “Diffusion-based methods propagate neighboring information into the missing regions [4, 2].”
Free-Form Image Inpainting With Gated Convolution (2018, background): “Moreover, inpainting results are improved by matching local features like image gradients [Ballester et al. 2001; Darabi et al. 2012] and statistics of similar patch offsets [He and Sun 2014].”
Generative Image Inpainting with Contextual Attention (2018, background): “Traditional diffusion or patch-based approaches such as [2, 4, 9, 10] typically use variational algorithms or patch similarity to propagate information from the background regions to the holes.”
Image Inpainting for Irregular Holes Using Partial Convolutions (2018, background): “Non-learning approaches to image inpainting rely on propagating appearance information from neighboring pixels to the target region using some mechanisms like distance field[3, 1, 30].”
Pluralistic Image Completion (2019, methodology): “Intra-Image Completion Traditional intra-image completion, such as diffusion-based methods [4, 1, 22] and patchbased methods [5, 7, 8, 3], assume image holes share similar content to visible regions; thus they would directly match, copy and realign the background patches to complete the holes.”
Semantic Image Inpainting with Progressive Generative Networks (2018, background): “However, as it is able to help propagate image intensity smoothly, it is also widely used in image inpainting task [1, 28, 32].”
Navier-stokes, fluid dynamics, and image and video inpainting (2001, methodology): “In order to obtain such a smooth interpolation and continuation of isophotes, it is necessary to go into high-order Partial Differential Equations (PDE’s) or systems of PDE’s, as done in [1, 2] and here (the authors of [7] have also recently introduced higher order models).”
Contrastive Learning for Diverse Disentangled Foreground Generation (2022, methodology): “Traditional methods [2,4,43,12,3] typically rely on low-level assumptions and image statistics, leading to over smoothing and results with limited visual semantics.”
Simultaneous structure and texture image inpainting (2003, methodology): “For image inpainting alternatives to this approach, see [ 2 ], [3].”
Text-Guided Neural Image Inpainting (2020, background): “Image inpainting Traditional diffusion-based or patch-based approaches [27, 31] fill missing regions by propagating neighbor information[2, 37] or copying from similar patches [8, 36] of the background based on low-level features.”
Image Fine-grained Inpainting (2020, neutral): “Traditional diffusion-based methods [3, 1] propagate information from neighboring regions to the holes.”
Boundless: Generative Adversarial Networks for Image Extension (2019, methodology): “Classical methods, such as [6, 8, 13, 12] typically rely on patch similarity and diffusion to borrow information from the known regions of the image to ﬁll in the hole.”
SPG-Net: Segmentation Prediction and Guidance Network for Image Inpainting (2018, background): “Traditional image inpainting approaches make use of the image-level features to propagate the texture from the surrounding context to the missing hole [3, 5].”
Nonlocal Linear Image Regularization and Supervised Segmentation (2007, background): “For example, a nonlocal generalization of variational inpainting [4] can be suggested.”
Non-local Regularization of Inverse Problems (2008, methodology): “Classical methods for inpainting use partial differential equations that propagate the information from the boundary of Ω to its interior, see for instance [36,37,38,39].”