COrigami - an AI-based Origami Design Tool

Images are licensed under the Creative Commons Attribution-NonCommercial 4.0 International License

I’m very excited to announce a paper I have co-authored: COrigami: An AI Pipeline for Co-Designing Flat-Foldable Visually Recognisable Origami. It introduces COrigami, a first of its kind AI-based origami design tool, capable of generating origami crease patterns and simulating their folded look, including simple shaping, starting from a textual description.

The majority of the ML research was carried out by a team at Google DeepMind, led by Tom Zahavy, with me and a number of other origami designers working mostly as domain experts. Nonetheless, the general approach of first using an LLM for translating a textual description of the model into a stick figure and then using deterministic algorithms to generate a crease pattern, is exactly the same as I suggested on my blog in 2023. I’m proud to have been able to contribute a little to this unique project, and to work with a great team of experts from two seemingly very different fields.

The image shows the progressive design stages from natural language to physical paper model. The workflow transitions from semantic stick figure generation and automated box-pleated crease pattern layout, through simulation and folding, to aesthetic shaping — ultimately establishing a viable structural starting point for physical artistic execution by Brandon Wong.