WIN Colloquium Series
Ye-Ji Mun
Shaping the Future of Coexistence: Systematic Inference and Influential Strategies for Safe and Effective Human-Robot Interaction
Wednesday, October 11th, 2023
12:00 PM - 1:00 PM
Unity Hall Rm 520
Abstract: Robots will soon be ubiquitous in everyday life—from autonomous vehicles in cities, assistive robots in homes, and collaborative manipulators in factories. To reach their full potential and ensure safe, effective interaction, such robots require intricate safety protocols and the capability to understand humans around them. In this presentation, I will present our efforts on using systematic inference and incorporating dynamic interaction strategies for designing robots that effectively collaborate with human teammates. I will introduce how structured inference enhances the accuracy of intention and objective estimation, subsequently facilitating seamless transitions between different types of interaction. Additionally, I will delve into how robot actions can influence and shape human behaviors towards desired goals and how human’s strategies change (perhaps antagonistically) in response as they repeatedly interact with the robot. I will illustrate how counterintuitive strategies that encourage unpredictable robot behavior can help maintain influence over time. This presentation aims to decode the complex interactions that occur when humans and robots coexist, providing insights into systematic inference and strategic robotic behavior to shape human behavior and uphold influential relationships.
Bio: Ye-Ji Mun is a PhD student in Human-Centered Autonomy Lab in the University of Illinois, Urbana-Champaign. Prior to her doctoral pursuit, she obtained the Bachelor’s degree in Electronics Engineering at Ewha Womans University, South Korea, in 2018. Her work stands at the intersection of robotics, computer vision, and artificial intelligence. Specifically, Ye-Ji studies human-robot interaction, concentrating on how autonomous agents and humans can achieve their shared or respective goals through safe and seamless interactions. Her work has been supported by industry partners (Foxconn Interconnect Technology, Ford) and the National Science Foundation. Ye-Ji is a recipient of the scholarship from Kwanjeong Educational Foundation.