Changwoo Yoo

I'm an undergraduate student at Korea University in Seoul, Korea. I'm currently working at KAIST AIPR Lab advised by Prof. Kee-Eung Kim. Before joining KAIST AIPR Lab, I worked at RLLAB advised by Prof. Youngwoon Lee and RILAB advised by Prof. Kyungjae Lee

Email  /  CV  /  Scholar  /  Github  /  LinkedIn

Research

My main research question is: "how to build generalist robots that can be applied to any tasks?". I believe offline reinforcement learning with large-scale data is a promising approach, but many real-world robotic datasets lack reward annotations. To address this, I am interested in reward-free learning from offline data, especially through offline and offline-to-online RL, unsupervised RL, self-supervised RL, and world models.

Learning a density-based reward from positive and negative demonstrations augmented with rule-based safety constraints to balance adaptability and reliability.