Rethinking Style Transfer: From Pixels to Parameterized Brushstrokes

TL;DR

This paper introduces a novel neural style transfer method that uses parameterized brushstrokes with differentiable rendering, leading to more natural stylizations and enhanced user control over the artistic process.

Contribution

It proposes a new parameterized brushstroke representation and a differentiable rendering mechanism for improved neural style transfer, moving beyond pixel-based methods.

Findings

Significantly improves visual quality of stylized images

Enables user control over brushstroke flow

Qualitative and quantitative evaluations confirm effectiveness

Abstract

There have been many successful implementations of neural style transfer in recent years. In most of these works, the stylization process is confined to the pixel domain. However, we argue that this representation is unnatural because paintings usually consist of brushstrokes rather than pixels. We propose a method to stylize images by optimizing parameterized brushstrokes instead of pixels and further introduce a simple differentiable rendering mechanism. Our approach significantly improves visual quality and enables additional control over the stylization process such as controlling the flow of brushstrokes through user input. We provide qualitative and quantitative evaluations that show the efficacy of the proposed parameterized representation.