Macrocanonical models for texture synthesis


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.

Associated communication

Arthur Leclaire presented this work during "The mathematics of imaging" 2019 trimester at Institut Henri Poincaré.

(don't miss the awesome intro)
I also presented this work at the 2019 SMAI colloquium

at Scale Space and Variational Methods in Computer Vision