Geometric styles transferred beyond original shape families
Open-Form Geometric Style Transfer
Computer Vision and Pattern RecognitionGraphics
Summary
Usually, geometric styles learned from one set of shapes can only be applied to similar shapes. This paper explores how to take a style learned from Latin letters and apply it to very different shapes, like Greek letters or even human poses. The authors created a method that separates the style from the basic shape structure, allowing the style to be flexibly used on new kinds of forms. They also built a large collection of varied shapes to help teach their system about many possible structures it might see.
What this means in practice
- •For graphic designers: Automatically apply unique geometric letter styles to diverse design elements like logos or icons beyond the original letter set.
- •For animation teams: Use learned geometric styles to stylize poses or characters consistently even when shapes differ significantly from training data.
Authors
Adar Elad, Gal Metzer, Omer Dahari, Min Lu, Daniel Cohen-Or
Abstract
A geometric style learned from one family of shapes is typically applied to new shapes from the same family. We consider a more challenging setting: learning the geometric design language of a stylized Latin alphabet and applying it to new forms, from Greek glyphs to entirely different structures such as human poses. We introduce Open-Form Geometric Style Transfer, where the forms to which a learned geometric style can be applied are not restricted to the shape family in which that style was observed. This setting is deceptively challenging: the difficulty is not merely learning a style, but learning the style without learning the shape family in which it was observed. We represent target forms by structural sketches and disentangle structural adherence from style realization using a Geometry LoRA and a style-specific LoRA. Critically, open-form transfer requires the Geometry LoRA to be universal. To achieve this, we devise a geometry-enrichment strategy that constructs a large and purposefully diverse collection of synthetic structural sketches spanning a broad range of unfamiliar geometries. Our experiments and ablations demonstrate the importance of both disentanglement and this universal geometry prior, enabling geometric styles learned from glyphs, pictograms, and other shape families to be transferred to substantially different, unseen forms.