Abstract:The goal of example-based texture synthesis methods is to generate arbitrarily large textures from limited exemplars in order to fit the exact dimensions and resolution required for a specific modeling task. The challenge is to faithfully capture all of the visual characteristics of the exemplar texture, without introducing obvious repetitions or unnatural looking visual elements. While existing non-parametric synthesis methods have made remarkable progress towards this goal, most such methods have been demonstrated only on relatively low-resolution exemplars. Real-world high resolution textures often contain texture details at multiple scales, which these methods have difficulty reproducing faithfully. In this work, we present a new general-purpose and fully automatic self-tuning non-parametric texture synthesis method that extends Texture Optimization by introducing several key improvements that result in superior synthesis ability. Our method is able to self-tune its various parameters and weights and focuses on addressing three challenging aspects of texture synthesis: i irregular large scale structures are faithfully reproduced through the use of automatically generated and weighted guidance channels; ii repetition and smoothing of texture patches is avoided by new spatial uniformity constraints; iii a smart initialization strategy is used in order to improve the synthesis of regular and near-regular textures, without affecting textures that do not exhibit regularities. We demonstrate the versatility and robustness of our completely automatic approach on a variety of challenging high-resolution texture exemplars.

Fast Texture Synthesis for Discrete Example-Based Elements

Fast Texture Synthesis Using Multiple Seeds As Constraints

Texture synthesis based on minimum energy cut and its applications

Fast Example-Based Surface Texture Synthesis Via Discrete Optimization.

Fast Texture Synthesis Based on Texture Standard Property

Optimization Technique of 2D Texture Synthesis

Fast Texture Synthesis by Feature Matching

Make-A-Texture: Fast Shape-Aware Texture Generation in 3 Seconds

A Fast Texture-by-numbers Synthesis Method Based on Texture Optimization

An Improved Example-Based Texture Synthesis Algorithm

A New Method for Texture Classification and Its Application in Texture Synthesis

Self Tuning Texture Optimization

An Efficient Method of Texture Synthesis Based on Graph Cuts

Appearance and geometry completion with constrained texture synthesis

Random Search Based Large Scale Texture Synthesis

A Texture Synthesis Method Based on Adaptive Patches

Efficient Solid Texture Synthesis Using Gradient Solids

A Rapid Texture Synthesis Algorithm Based On Clustering Preprocessing

Inverse Texture Synthesis

Fast near-regular texture synthesis based on the concept of co-evolution

Non-Local Texture Optimization With Wasserstein Regularization Under Convolutional Neural Network