Conference

ICCC 2020

Authors

• Devi Parikh
• C. Lawrence Zitnick

Contributions

• The roles of creating stroke proposals and voting could each be fulfilled by either humans (H) or machines (M). Borrowing terminology from Generative Adversarial Networks we can call the former role a generator (G), and the latter a discriminator (D). This allows for 4 {H,M} × {G,D} co-creation scenarios. Further, different individuals could play the role of generators/discriminators across iterations, leading to crowd co-creation.
• In this work, as a step towards human-machine cocreation, a study is made on various human-human crowd co-creation scenarios. In the first, Individual, a single human creates the entire sketch (no discriminator D, and no crowd). Second, in Collaborative the sketch is generated by multiple human agents (crowd) iteratively taking turns adding strokes. That is, all the agents act as generators G and there is no voting or discriminator D. The third, Collaborative + voting, is where multiple human agents (generators) propose new strokes at each iteration. Another set of human agents (discriminators) vote on which set of strokes to add to the sketch. Finally, explored Individual with collaborative prompts, for which the crowd is involved indirectly. A single human creates the entire sketch, but by following text prompts that describe the evolution of a sketch that was created in the Collaborative scenario.

Approach

• Created sketch interface.
• Explored all these above creation steps.

Result

• Measured across 4 axes.
• Individual is rated good quality but poor novelty.
• Collaborative works good novelty but poor quality.
• Individual with collaborative prompts does poor.
• Collaborative with voting is good quality and good novelty.