Macrocanonical Models for Texture Synthesis
This work addresses texture synthesis for computer vision applications, but it is incremental as it builds on existing macrocanonical models and focuses on theoretical extensions.
The paper tackles the problem of extending macrocanonical models for texture synthesis from quantized to real-valued images, establishing conditions for this extension and analyzing an algorithm that alternates between sampling and minimization, with experiments using neural network features.
In this article we consider macrocanonical models for texture synthesis. In these models samples are generated given an input texture image and a set of features which should be matched in expectation. It is known that if the images are quantized, macrocanonical models are given by Gibbs measures, using the maximum entropy principle. We study conditions under which this result extends to real-valued images. If these conditions hold, finding a macrocanonical model amounts to minimizing a convex function and sampling from an associated Gibbs measure. We analyze an algorithm which alternates between sampling and minimizing. We present experiments with neural network features and study the drawbacks and advantages of using this sampling scheme.