Long Range Constraints for Neural Texture Synthesis Using Sliced Wasserstein Loss

TL;DR

This paper introduces a novel multi-scale texture synthesis method using Sliced Wasserstein Loss that captures long-range constraints without requiring user-provided spatial tags, improving over existing exemplar-based algorithms.

Contribution

The authors propose a new set of statistics based on Sliced Wasserstein Loss and a multi-scale approach for texture synthesis that does not need spatial tags, addressing limitations of prior methods.

Findings

Effective long-range constraint capture in textures

Comparable or improved synthesis quality without spatial tags

Outperforms some existing optimization-based algorithms

Abstract

In the past decade, exemplar-based texture synthesis algorithms have seen strong gains in performance by matching statistics of deep convolutional neural networks. However, these algorithms require regularization terms or user-added spatial tags to capture long range constraints in images. Having access to a user-added spatial tag for all situations is not always feasible, and regularization terms can be difficult to tune. Thus, we propose a new set of statistics for texture synthesis based on Sliced Wasserstein Loss, create a multi-scale method to synthesize textures without a user-added spatial tag, study the ability of our proposed method to capture long range constraints, and compare our results to other optimization-based, single texture synthesis algorithms.