Skip to content

Latest commit

 

History

History
20 lines (12 loc) · 1.42 KB

README.md

File metadata and controls

20 lines (12 loc) · 1.42 KB

QRt: QR Code Art Generator

Project description

Research in the field of QR code art generation has gained notable attention in recent years. The fusion of technology and creativity opens up new possibilities beyond mere aesthetics. However, making such QR codes both pleasing for humans and readable by modern devices is a complex task requiring both a high-quality generative model and a QR parser for evaluation. In this project, we discover methods to make such a pipeline possible and generate both stable and artful QR codes with little-to-no ambiguity.

As a result of the project, we provide a pipeline to generate visually appealing QR codes with the popular diffusion model Stable Diffusion followed by ControlNets.

pipeline

Links to workflow

If you are interested in this project, you can read our research. In addition, please check the following notebooks: Inference Notebook and Evaluation Notebook.

Credits

Project developed and done by:

  • Polina Zelenskaya, Innopolis University
  • Karina Denisova, Innopolis University
  • Leila Khaertdinova, Innopolis University