Texture synthesis through convolutional neural networks and spectrum constraints

This paper presents a significant improvement for the synthesis of texture images using convolutional neural networks (CNNs), making use of constraints on the Fourier spectrum of the results. More precisely, the texture synthesis is regarded as a constrained optimization problem, with constraints conditioning both the Fourier spectrum and statistical features learned by CNNs. In contrast with existing methods, the presented method inherits from previous CNN approaches the ability to depict local structures and fine scale details, and at the same time yields coherent large scale structures, even in the case of quasi-periodic images. This is done at no extra computational cost. Synthesis experiments on various images show a clear improvement compared to a recent state-of-the art method relying on CNN constraints only.

• Publication year

  2016

• Venue

  International Conference on Pattern Recognition

• Publication date

  2016-05-04

• Fields of study

  Computer Science, Engineering

• Identifiers
  DOI 10.1109/ICPR.2016.7900133arXiv 1605.01141
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Exploring generative perspective of convolutional neural networks by learning random field models

  2018cited by this paper
• Texture Synthesis Using Shallow Convolutional Networks with Random Filters

  2016cited by this paper
• Texture Networks: Feed-forward Synthesis of Textures and Stylized Images

  2016cited by this paper
• Conditional Gaussian Models for Texture Synthesis

  2015cited by this paper
• Texture Synthesis Using Convolutional Neural Networks

  2015influential reference
• Deep filter banks for texture recognition and segmentation

  2015cited by this paper
• Variational Texture Synthesis with Sparsity and Spectrum Constraints

  2015cited by this paper
• Generative Modeling of Convolutional Neural Networks

  2014cited by this paper
• Very Deep Convolutional Networks for Large-Scale Image Recognition

  2014cited by this paper
• Synthesizing and Mixing Stationary Gaussian Texture Models

  2014cited by this paper
• Return of the Devil in the Details: Delving Deep into Convolutional Nets

  2014cited by this paper
• Exemplar-based Texture Synthesis: the Efros-Leung Algorithm

  2013cited by this paper
• Micro-Texture Synthesis by Phase Randomization

  2011cited by this paper
• Random Phase Textures: Theory and Synthesis

  2011cited by this paper
• Parallel controllable texture synthesis

  2005cited by this paper
• A Parametric Texture Model Based on Joint Statistics of Complex Wavelet Coefficients

  2000cited by this paper
• Fast texture synthesis using tree-structured vector quantization

  2000cited by this paper
• Texture synthesis by non-parametric sampling

  1999cited by this paper
• Algorithm 778: L-BFGS-B: Fortran subroutines for large-scale bound-constrained optimization

  1997cited by this paper
• Pyramid-based texture analysis/synthesis

  1995cited by this paper
• Markov Random Field Texture Models

  1983cited by this paper
• Visual Pattern Discrimination

  1962cited by this paper

• TiPGAN: High-quality tileable textures synthesis with intrinsic priors for cloth digitization applications

  2025cites this paper
• Deep SVBRDF Acquisition and Modelling: A Survey

  2024cites this paper
• Wasserstein Distortion: Unifying Fidelity and Realism

  2023cites this paper
• A Neurally Guided Patch-Based Style Transfer for Mobile Devices

  2023cites this paper
• The infinite doodler: expanding textures within tightly constrained manifolds

  2023cites this paper
• Training StyleGAN2 on BWW-Texture Dataset: Generating Textures Through Transfer Learning

  2023cites this paper
• Learning in a Single Domain for Non-Stationary Multi-Texture Synthesis

  2023cites this paper
• Finding Biological Plausibility for Adversarially Robust Features via Metameric Tasks

  2022cites this paper
• Long Range Constraints for Neural Texture Synthesis Using Sliced Wasserstein Loss

  2022cites this paper
• Development of Texture Mapping Approaches for Additively Manufacturable Surfaces

  2022cites this paper
• Generalized Rectifier Wavelet Covariance Models For Texture Synthesis

  2022cites this paper
• Controlled Synthesis of Fibre-reinforced Plastics Images from Segmentation Maps using Generative Adversarial Neural Networks

  2022cites this paper
• A Convolutional Neural Networks-Based Approach for Texture Directionality Detection

  2022cites this paper
• Reconstructing decorative texture of wooden board based on modified image quilting

  2021cites this paper
• Structure Preserving Exemplar-Based 3D Texture Synthesis

  2021cites this paper
• TGHop: an explainable, efficient, and lightweight method for texture generation

  2021cites this paper
• Cyclostationary Gaussian noise: theory and synthesis

  2021cites this paper
• Generative modelling of BRDF textures from flash images

  2021cites this paper
• Neural FFTs for Universal Texture Image Synthesis

  2020cites this paper
• Machine Learning Methods for Reverse Engineering of Defective Structured Surfaces

  2020cites this paper
• Co-occurrence based texture synthesis

  2020cites this paper
• Pitfalls of the Gram Loss for Neural Texture Synthesis in Light of Deep Feature Histograms

  2020influential citation
• Wasserstein Generative Models for Patch-based Texture Synthesis

  2020cites this paper
• A Stochastic Multi-layer Algorithm for Semi-discrete Optimal Transport with Applications to Texture Synthesis and Style Transfer

  2020cites this paper
• Transposer: Universal Texture Synthesis Using Feature Maps as Transposed Convolution Filter

  2020cites this paper
• High-Resolution Neural Texture Synthesis with Long-Range Constraints

  2020influential citation
• NITES: A Non-Parametric Interpretable Texture Synthesis Method

  2020cites this paper
• A Sliced Wasserstein Loss for Neural Texture Synthesis

  2020influential citation
• Conditional Generative ConvNets for Exemplar-Based Texture Synthesis

  2019cites this paper
• Modélisation de fonds complexes statiques et en mouvement : application à la détection d'événements rares dans les séries d'images

  2019cites this paper
• Layered leaf texturing using structure-guided model

  2019cites this paper
• Macrocanonical Models for Texture Synthesis

  2019cites this paper
• On Demand Solid Texture Synthesis Using Deep 3D Networks

  2019cites this paper
• Non-Local Texture Optimization With Wasserstein Regularization Under Convolutional Neural Network

  2019cites this paper
• Perception Evaluation: A New Solar Image Quality Metric Based on the Multi-fractal Property of Texture Features

  2019cites this paper
• A deep-level region-based visual representation architecture for detecting strawberry flowers in an outdoor field

  2019cites this paper
• Thoracic MRI Emulation Through Texture Synthesis and Ground Truth Manipulation

  2019cites this paper
• Statistical Texture Features

  2019cites this paper
• Maximum entropy methods for texture synthesis: theory and practice

  2019cites this paper
• Texture synthesis with neural networks

  2018influential citation
• Detection of Strawberry Flowers in Outdoor Field by Deep Neural Network

  2018cites this paper
• Texture Mixing by Interpolating Deep Statistics via Gaussian Models

  2018influential citation
• One-shot Texture Segmentation

  2018cites this paper
• A Texture Synthesis Model Based on Semi-Discrete Optimal Transport in Patch Space

  2018cites this paper
• Repetitiveness Metric of Exemplar for Texture Synthesis

  2017cites this paper
• Texture and art with deep neural networks.

  2017cites this paper
• A survey of exemplar-based texture synthesis

  2017cites this paper
• A parametric texture model based on deep convolutional features closely matches texture appearance for humans

  2017influential citation
• Deep correlations for texture synthesis

  2017cites this paper
• What does it take to generate natural textures?

  2017cites this paper
• Image Inpainting for High-Resolution Textures using CNN Texture Synthesis

  2017cites this paper
• Deep Correlations for Texture Synthesis

  2017influential citation
• A Novel Texture Exemplars Extraction Approach Based on Patches Homogeneity and Defect Detection

  2017cites this paper
• Stationary dynamic texture synthesis using convolutional neural networks

  2016cites this paper
• Texture synthesis using convolutional neural networks with long-range consistency and spectral constraints

  2016influential citation
• Universal Semantic-less Texture Boundary Detection for Microscopy (and Metallography)

  year unknowncites this paper