On pixel-based texture synthesis by non-parametric sampling

By Seunghyup Shin, Tomoyuki Nishita, Sung Yong Shin. Computers & graphics, 2006

695 citations recorded by the source corpus.

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

View Paper: https://www.semanticscholar.org/paper/a6aa4552f2187447853425d0b91a94b19f2f3059

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

Citations in the corpus

Kernel regression estimation for continuous spatial processes (2007, methodology): “The method of Efros and Leung [ 22 ] is based on an heuristic texture synthesis method while that of Levina and Bickel [31] is a nonparametric bootstrap of stationary random fields.”
Learning Pyramid-Context Encoder Network for High-Quality Image Inpainting (2019, methodology): “Patchbased methods were first proposed for texture synthesis [6, 7].”
3D Photography Using Context-Aware Layered Depth Inpainting (2020, methodology): “Inspired by the success of texture synthesis [9, 8], examplebased methods complete the missing regions by transferring the contents from the known regions of the image, either through non-parametric patch-based synthesis [58, 1, 5, 20] or solving a Markov Random Field model using belief propagation ”
N-dimensional probability density function transfer and its application to color transfer (2005, background): “The recent breakthrough in texture synthesis [ 4 ] is probably one of the most significant examples of this idea.”
Non-metric affinity propagation for unsupervised image categorization (2007, methodology): “Exemplars have been used with success in a variety of vision tasks, including image synthesis [3, 4 ], super-resolution [5, 6], image and video completion [7, 8] and combinedy tracking and object detection [9, 10].”
Fast image and video denoising via nonlocal means of similar neighborhoods (2005, methodology): “This new concept for image denoising is popular in other image processing areas, such as texture synthesis, where a new pixel is synthesized as the weighted average of known image pixels with similar neighborhoods [ 5 ], [16], [18].”
Visualizing and Understanding Deep Texture Representations (2015, methodology): “For instance, Efros and Lueng [10] proposed a pixel-by-pixel synthesis approach based on sampling similar patches – the method was simple and effective for a wide range of textures.”
Context as Supervisory Signal: Discovering Objects with Predictable Context (2014, background): “That is, we “grow” the predicted region one small step at a time, reminiscent of texture synthesis [27].”
Mask-Specific Inpainting with Deep Neural Networks (2014, methodology): “Many successful methods (some based on texture synthesis [8]) have been proposed in the past, for instance [7, 17, 5, 26, 13] and references therein.”
“Double-DIP”: Unsupervised Image Decomposition via Coupled Deep-Image-Priors (2018, background): “This strong internal patch recurrence was exploited for solving a large variety of computer vision tasks [9, 13, 16, 15, 19, 36, 29, 7, 11].”
Interactive image completion with perspective correction (2006, methodology): “Example-based approaches have their origin in texture synthesis, where large texture patches are synthesized from small texture samples on a per-pixel basis using a pixel neighborhood for color prediction [2, 14, 32].”
The Synthesizability of Texture Examples (2014, background): “The third group generate textures by copying pixels or patches from the exemplar inputs [8, 7, 16, 15, 5].”
Learning generative models of mid-level structure in natural images (2012, methodology): “In order to make progress towards our goal of developing more structured models of natural images we then devise a modification of the FoE and demonstrate how changing the structure of the model and using a more powerful learning algorithm can substantially increase the generative power, giving rise”