Drawcto

Drawcto is an AI that does turn-taking non-representational drawing with humans.

Non-representational art, such as works by Wassily Kandinsky, Joan Mitchell, Willem de Kooning, etc., showcases diverse artistic expressions and challenges to viewers with its interpretive open-ended-ness and lack of a clear mapping to our everyday reality. Human cognition and perception nonetheless aid us in making sense of, reasoning about, and discussing the perceptual features prevalent in such non-representational art. While there have been various Computational Creative systems capable of generating representational artwork, only a few existing Computational (Co)Creative systems for visual arts can produce non-representational art. Inspired by this, we propose a new drawing application Drawcto, a multi-agent AI that incorporates elements of the human visual perception theory capable of co-creating non-representational art with a human collaborator on the web.

This work is selected into the Experimental AI in Games workshop at the 17th AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE-21).

Personal Contribution

This work is a collaboration of Manoj Deshpande, Arpit Mathur, Bhavika Devnani, Laney Light, Luowen Qiao, and myself, and it is supervised by Dr. Brian Magerko. My personal contribution are listed as the following:

• Designed the multi-agent architecture

• Designed and implemented the shape-grammar agent, which reacts to user strokes based on theories of shape-grammar.

• Implemented the backend web application.