With the rise in the use of artificial intelligence (AI), AI-driven systems are being increasingly placed at homes, hospitals, offices, manufacturing and other places to support people. To achieve desired goals, improve human experience and enhance human-AI collaboration efficiency, it is important to understand the collaborative processes between human and AI agents during the process of achieving shared goals.
This project involves designing, implementing and evaluating several collaborative scenarios based on an existing game, Overcooked-AI (https://github.com/HumanCompatibleAI/overcooked_ai). More specifically, this may include
designing several collaborative scenarios for 1) human-human, 2) human-AI to compare and evaluate human experience during the collaboration,
modifying the current Overcooked-AI game if necessary,
testing different AI implementation and their performance
conducting a user study to collect participants’ experience with the game via surveys, interviews, and/or physiological signals.
analysing the collected responses.
If you are interested in participating in this project, please email Priscilla (firstname.lastname@example.org) and Xuanying (email@example.com) with your academic transcript(s), CV, and a few sentences about why you are interested in this project.
Supervisors for this project are:
Priscilla Kan John (School of Computing)
Xuanying Zhu (School of Cybernetics)
Paul Wong (School of Cybernetics)
Completed coursework or equivalent experience in programming
preferably has done the Human Computer Interaction course (COMP3900/6390)
self-motivated and ability to work independently with provided guidance
A genuine interest in your project topic and desire to learn new skills
Human-AI interaction, Human-computer interaction, biometrics, Artificial Intelligence