*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Zhiyu Lin
PhD Student in Computer Science
School of Interactive Computing, College of Computing
Georgia Institute of Technology
Committee
Dr. Mark Riedl (Advisor, School of Interactive Computing, Georgia Institute of Technology)
Dr. Alan Ritter (School of Interactive Computing, Georgia Institute of Technology)
Dr. Gil Weinburg (School of Music, Georgia Institute of Technology)
Dr. Matthew Guzdial (Department of Computing Science, University of Alberta)
Dr. Wei Xu (School of Interactive Computing, Georgia Institute of Technology)
Abstract
Recent advancements in controlling generative systems allowed more fine-grained control allowing automation of more parts of game development and creative content creation, several problems arise preventing human creators from fully utilizing these techniques, including challenges in high-level intent-based control, lack of intuitiveness, difficulty in personalization, AI-centric interfaces for human creators, and ultimately not aiming at enhancing the creativity of human designers in a collaborative manner.
To solve these problems, I purpose Human-Aware Artificial Intelligence Procedural Content Generation (HAAI-PCG), a mixed-initiative co-creativity system with human awareness, to overcome these problems.
I posit that such a system will empower high-level intent-based customization of contents and use the capability of high-level control to enable natural personalized collaborative interaction between a human designer and AI, to better inspire human designers to create content satisfied to both humans and AI than current PCG systems.
I have previously discovered how a PCG-Machine-Learning (PCGML) system can utilize high-level intents and how an AI agent take advice from human to learn faster and perform better;
I propose to study how a combination of these techniques will fuel a co-creativity system that builds the interactions between human designers and AI agents, by investigating (1) preferred interactions between participants of the system, (2) quality of co-created contents from this system, and (3) how the system will be perceived by human designers, the user of this system.
The investigation will open the door to the key understanding of major features in a human-aware co-creativity system, how human designers can benefit from it, and ultimately, the potential of co-creative systems in this new generation of AI and ML.
