Skip to content

Out-of-control package that continuously generate new graffiti artworks based on feed image data

License

Notifications You must be signed in to change notification settings

graffiti-ai/learned-styles

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Graffiti AI: Generative Adversarial Style Transfer for Artists

Abstract

This repository presents Graffiti AI, an approach to artistic style transfer using generative adversarial networks (GANs). Graffiti AI is a collaborative research project between Andrew Stepin and street art and graffiti artist Linelogic. The goal of the project is to develop a GAN model capable of learning artist distinctive style from his existing artworks and sketches, and subsequently generating new unique images that capture the essence of his style. The proposed method leverages the power of adversarial training to separate and recombine the content and style representations of images, enabling the generation of new artworks that preserve the artist's unique visual aesthetics while introducing novel content.

Taining timelapse of the first model: https://www.youtube.com/watch?v=J_i_dvPpku4

Releases

No releases published

Packages

No packages published