Mon, Oct 02, 2023 @ 11:00 AM - 12:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Dr. Erdem Biyik, Assistant Professor, USC Thomas Lord Department of Computer Science
Talk Title: Efficient Preference Learning for Interactive Robots
Abstract: In the realm of human-robot interaction and multi-agent systems, decentralized agents often collaborate to achieve a common task. The crux of effective collaboration lies in the ability to predict and anticipate the actions of other agents. While we, humans, excel at this due to our innate capacity to understand intentions, robots currently lag behind. This talk delves into the integration of reward learning and partner modeling as a way to address this challenge. I will present active learning techniques tailored to this challenge and discuss the pivotal role of preference data for learning rewards in the settings where demonstration data may fail. Finally, I am going to show some preliminary results and examples of how partner-modeling enables decentralized agents to cooperate efficiently.
Biography: Erdem Biyik is an assistant professor in Thomas Lord Department of Computer Science at the University of Southern California. He leads the Learning and Interactive Robot Autonomy Lab (Lira Lab). Prior to joining USC, he was a postdoctoral researcher at UC Berkeley's Center for Human-Compatible Artificial Intelligence. He received his Ph.D. and M.Sc. degrees in Electrical Engineering from Stanford University, working at the Stanford Artificial Intelligence Lab (SAIL), and his B.Sc. degree in Electrical and Electronics Engineering from Bilkent University in Ankara, TÃ¼rkiye. During his studies, he worked at the research departments of Google and Aselsan. Erdem was an HRI 2022 Pioneer and received an honorable mention award for his work at HRI 2020. His works were published at premier robotics and artificial intelligence journals and conferences, such as IJRR, CoRL, RSS, NeurIPS.
Host: Dr. Somil Bansal, email@example.com
Audiences: Everyone Is Invited
Contact: Mayumi Thrasher