Texture optimization for example-based synthesis

By Vivek Kwatra, Irfan Essa, Aaron F. Bobick, Nipun Kwatra. 2005

770 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition, Generative Models

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 Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

In Generative Models

Citations in the corpus

Regionwise Generative Adversarial Image Inpainting for Large Missing Areas (2019, neutral): “Patch-based methods [20], [21], on the other hand, fill in missing regions with patches from a collection of source images that maximize the patch similarity.”
Free-Form Image Inpainting With Gated Convolution (2018, methodology): “…2001; Barnes et al. 2009; Drori et al. 2003; Sun et al. 2005], structures [Huang et al. 2014], transformation or distortion information [Huang et al. 2013; Pavi´c et al. 2006] and image exemplars [Criminisi et al. 2004; Hays and Efros 2007; Kwatra et al. 2005; Whyte et al. 2009; Zhao et al. 2018].”
A Learned Representation For Artistic Style (2016, methodology): “Kwatra et al. (2005) approaches the texture synthesis problem from an energy minimization perspective, progressively reﬁning the texture using an EM-like algorithm.”
Precomputed Real-Time Texture Synthesis with Markovian Generative Adversarial Networks (2016, methodology): “In order to have a coherent transition between adjacent patches, we blend their output gradient like texture optimization [18] did.”
Combining Markov Random Fields and Convolutional Neural Networks for Image Synthesis (2016, methodology): “Multi-resolution synthesis [12, 14, 26] provides some improvement here (and we adapt this in our method, too), but a principled solution requires additional high-level constraints.”
Image Inpainting for Irregular Holes Using Partial Convolutions (2018, methodology): “Patch-based methods such as [5,15] operate by searching for relevant patches from the image’s non-hole regions or other source images in an iterative fashion.”
High-Resolution Image Inpainting Using Multi-scale Neural Patch Synthesis (2016, background): “The first group of approaches relies on texture synthesis techniques, which fills in the hole by extending textures from surrounding regions [14, 13, 27, 26, 6, 12, 40, 41, 23, 24, 2].”
Fast Patch-based Style Transfer of Arbitrary Style (2016, background): “Early works deﬁne style using similarity measures or local statistics based on the pixel values [7, 8, 13, 14, 20, 24].”
Image Completion Using Global Optimization (2006, methodology): “Some methods try to alleviate this problem by taking a more global approach and using an EM-like scheme for optimizing the completion process [8, 13].”
SPG-Net: Segmentation Prediction and Guidance Network for Image Inpainting (2018, methodology): “Later works using patch-based methods [4, 19] could optimize the inpainting performance by searching the best matching patches.”
Globally and locally consistent image completion (2017, methodology): “This was later extended with image stitching [Kwatra et al. 2003] with graph cuts and texture generation [Kwatra et al. 2005] based on energy optimization.”
Transformation guided image completion (2013, background): “, coordinate descent) [25, 15] or MRF models [14, 13] which can be solved efficiently using belief propagation.”
Image upsampling via imposed edge statistics (2007, methodology): “On the other hand, example-based techniques have shown more success in synthesizing textures than parametric approaches, see [Kwatra et al. 2005].”
Learning a Neural 3D Texture Space From 2D Exemplars (2019, background): “Computer vision typically had looked into generating textures from exemplars, such as by non-parametric sampling [4], vector quantization [25], optimization [12] or nearestneighbor field synthesis (PatchMatch [2]) with applications in in-painting and also (3D) graphics.”
Regenerative morphing (2010, methodology): “Our optimization framework is inspired by recent example-based methods [17, 10, 16, 2] that showed impressive image and video synthesis results for related problems such as retargeting, reshuffling, holefilling and texture synthesis.”
Image Inpainting with Learnable Feature Imputation (2020, background): “For example, pattern matching, by searching and replacing with similar patches [4,8,23,29,38,43], or diffusion-based, by smoothly propagating information from the boundary of the missing area [3,5,6].”
Text-Guided Neural Image Inpainting (2020, background): “Some user-guided image inpainting approaches allow external guidance, such as a line [41] and edges [24], and exemplar [13].”
Image Compression With Edge-Based Inpainting (2007, background): “So far, attractive results have been achieved by newly presented texture synthesis techniques to generate regions of homogeneous textures from their surroundings [6]–[14].”