Biography
Currently, I am leading the perception and the learning tasks for the DARPA Aircrew-Labor In-Cockpit Automation System (ALIAS) program (co-PI) that aims to bring intelligence to cockpits through semantic perception and learning new skills from observing experienced pilots. I am also a PI on the intelligence architecture subtask of the ARL RCTA program where my team’s work on language understanding on robot navigation won the Best Cognitive Robotics Paper Award at the IEEE International Conference on Robotics and Automation (ICRA) in 2015. My newest project is the US DoD – Korea MOTIE (co-PI) collaboration on robotics technologies for disaster response, where I will be leading the semantic map construction by leveraging text data from social media and crowdsourcing.
Areas of Expertise
Social
Education
Yonsei University
B.S.
Biotechnology
Columbia University
M.S.
Computer Science
Carnegie Mellon University
Ph.D.
Language and Information Technologies
Patents
Vehicle operator workload estimation system and method
12504820
2025-12-23
An estimation system includes a plurality of sensors that generate a multimodal signal, where the multimodal signal indicates a state of a user. The estimation system also includes at least one processor that receives the multimodal signal from the plurality of sensors, and determines a workload experienced by the user based on the multimodal signal and a workload model, wherein the workload model relates multimodal signal data to an experienced workload.
Articles
Safety Not Found (404): Hidden Risks of LLM-Based Robotics Decision Making
arXivJ Han, J Seo, J Min, J Oh, J Kim
2026-01-09
One mistake by an AI system in a safety-critical setting can cost lives. As Large Language Models (LLMs) become integral to robotics decision-making, the physical dimension of risk grows; a single wrong instruction can directly endanger human safety. This paper addresses the urgent need to systematically evaluate LLM performance in scenarios where even minor errors are catastrophic.
InterPReT: Interactive Policy Restructuring and Training Enable Effective Imitation Learning from Laypersons
arXivFG Zhu, J Oh, R Simmons
2026-02-04
Imitation learning has shown success in many tasks by learning from expert demonstrations. However, most existing work relies on large-scale demonstrations from technical professionals and close monitoring of the training process.
Jean Oh
Google ScholarFull publication list: https://scholar.google.com/citations?hl=en&user=jizMCFAAAAAJ&view_op=list_works&sortby=pubdate


