Collaborative Drawing Systems

Lead Research Organisation: King's College London
Department Name: Informatics

Abstract

There is an established research history of human-computer improvisation systems within music, performance, and jazz. However, an open research question is how a human and a computer would collaborate in an open manner within visual arts, in particular sketching and drawing. Faced with the proposition of such a collaborative sketching system, there is a lack of understanding what the clear use cases would be for how such a collaboration would provide utility for a visual artist. As a result, the principal research question for the thesis is how can technology and current developments in Artificial Intelligence and Computer Vision contribute to the drawing process for an artist through collaboration?

To begin to answer this question, one must understand the utility of drawing itself, and what it contributes to the creative process for practitioners such as professional illustrators, fine artists, and art students. One can address through a mixed-methods qualitative and quantitative research study. The aim being to discover possible roles that technology could play in observing, modelling, and possible assisting an artist with their drawing practise. The outcomes of this study will inform a set of experimental interactions within a modular real-time system that observes, learns from, and interacts with the visual artist. Through the development and study of such a proposed research system, one can answer and understand the principal research question mentioned above.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/N509498/1 01/10/2016 30/09/2021
1949876 Studentship EP/N509498/1 01/10/2017 30/03/2021 Chipp Jansen
 
Description Based on a small pilot study (Autumn 2018) of drawing practitioners (full-time professional illustrators, students on drawing courses, and part-time drawing enthusiasts), some themes on collaborating with technology have been discovered:
1. Physical mediums are still a traditional and initial way for artists to generate work
2. Collaborative drawing is uncommon in the workflows of drawing practitioners, in general. When asked about envisioning collaborating with an AI, a co-creative AI is preferred to a corrective or didactic AI.
3. Artists are wary of automation of creative work, which raises questions of authorship and employment viability.

This has informed the creation of a drawing data gathering prototype system which has been used in a drawing data gathering user study (Winter 2020) that has gathered a research data set of drawing activity from full-time drawing practitioners. It's anticipated that this research dataset will contribute towards:
1. Development of analytic research methodologies for drawing activity
2. Development of models of drawing to be experimented with creative AI
3. Contribution towards the development of a collaborative drawing system
Exploitation Route Academically it is envisioned that the research will contribute towards these areas:
1. Creative AI through the development of the drawing AI model
2. Human-robotic interaction and robotic systems research with drawing technologies
3. Computer vision with learning from video (of drawing activities)
4. Digital humanities research with outcomes from analysis of drawing activity

Non-academically, a collaborative drawing system might contribute towards the creative practice of artists to potentially push new creative workflows. It also contributes potentially toward the development of new creative technology tools and towards assistive technology for drawing practitioners with diminished drawing facility.
Sectors Creative Economy,Digital/Communication/Information Technologies (including Software),Electronics,Healthcare,Culture, Heritage, Museums and Collections

URL https://nms.kcl.ac.uk/chipp.jansen/research/drawing/overview/