Beomjoon Kim
{firstname}.{lastname} at kaist.ac.kr

I am an Assistant Professor in Graduate School of AI at KAIST. I direct the Intelligent mobile-manipulation (IM^2) lab, where we have several research internship positions available.

I am interested in creating general-purpose mobile manipulation robots that can efficiently make decisions in complex environments.

Previously, I obtained my Ph.D. in computer science from MIT CSAIL, my MSc in computer science from McGill University, and my BMath in computer science and statistics from University of Waterloo.

CV  /  Google Scholar  /  Github  /  Twitter

profile photo
Recent research

Check out my talk at MIT Embodied Intelligence seminar for our recent effort in developing a general-purpose robot.

IM^2 lab

Ph.D students

Jamie (Yoonyoung) Cho
Dongwon Son
Minchan Kim
Haewon Jung

Masters students

Jaehyung Kim
Dongryung Lee
Jisu Han
Junhyek Han
Sanghyeon Son
Sejune Joo

Interns

Hojin Jung
Dongwuk Lee

Publications

An Intuitive Multi-Frequency Feature Representation for SO(3)-Equivariant Networks
Dongwon Son, Jaehyung Kim, Sanghyeon Son, Beomjoon Kim
International Conference on Learning Representations (ICLR), 2024. [pdf]

Open X-Embodiment: robotic learning datasets and RT-X models
Open X-Embodiment Collaboration
International Conference on Robotics and Automation (ICRA), 2023. [pdf]

Learning whole-body manipulation for quadrupedal robot
Seunghun Jeon, Moonkyu Jung, Suyoung Choi, Beomjoon Kim (co-corresponding author), Jemin Hwangbo
Robotics and Automation Letters (RA-L), 2023. [pdf] [video]

Preference learning for guiding the tree search in continuous POMDPs
Jiyong Ahn, Sanghyeon Son, Dongryung Lee, Jisu Han, Dongwon Son, and Beomjoon Kim.
Conference on Robot Learnining (CoRL), 2023. [pdf] [project page]

Pre- and Post-Contact Policy Decomposition for Non-Prehensile Manipulation with Sim-to-Real Transfer.
Minchan Kim, Junhyek Han, Jaehyung Kim, and Beomjoon Kim.
International Conference on Intelligent Robots and Systems (IROS), 2023. [pdf] [project page]

Local object crop collision network for efficient simulation of non-convex objects in GPU-based simulators.
Dongwon Son, Beomjoon Kim.
Robotics: Science and Systems (RSS), 2023. [pdf] [project page]

Ohm^2: Optimal hierarchical planner for object search in large environments via mobile manipulation
Yoonyoung Cho*, Donghoon Shin*, Beomjoon Kim.
International Conference on Intelligent Robots and Systems (IROS), 2022. [pdf]

Representation, learning, and planning algorithms for geometric task and motion planning
Beomjoon Kim, Luke Shimanuki, Leslie Pack Kaelbling, Tomás Lozano-Pérez.
International Journal of Robotics Research, 2021. [doi] [arXiv]

Integrated task and motion planning .
Caelan Reed Garrett, Rohan Chitnis, Rachel Holladay, Beomjoon Kim, Tom Silver, Leslie Pack Kaelbling, Tomás Lozano-Pérez.
Annual Review of Control, Robotics, and Autonomous Systems, 2021. [pdf] [arXiv]

A Long Horizon Planning Framework for Manipulating Rigid Pointcloud Objects .
Anthony Simeonov, Yilun Du, Beomjoon Kim, Francois Hogan, Joshua Tenenbaum, Pulkit Agrawal, Alberto Rodriguez.
Conference on Robot Learning (CoRL), 2020. [arXiv] [project page] [video]

CAMPs: Learning Context-Specific Abstractions for Efficient Planning in Factored MDPs .
Rohan Chitnis*, Tom Silver*, Beomjoon Kim, Leslie Pack Kaelbling, Tomás Lozano-Pérez
Conference on Robot Learning (CoRL), 2020. Plenary talk (top 12% of accepted papers). [pdf] [video]

Monte Carlo Tree Search in continuous spaces using Voronoi optimistic optimization with regret bounds .
Beomjoon Kim, Kyungjae Lee, Sungbin Lim, Leslie Pack Kaelbling, Tomas Lozano-Perez.
AAAI Conference on Artificial Intelligence (AAAI), 2020. Oral (top 6% of accepted papers). [pdf] [appendix]

Learning to guide task and motion planning using score-space representation.
Beomjoon Kim, Zi Wang, Leslie Pack Kaelbling, Tomás Lozano-Pérez.
International Journal of Robotics Research, 2019. [doi] [arXiv]

Learning value functions with relational state representations for guiding task-and-motion planning.
Beomjoon Kim, Luke Shimanuki
Conference on Robot Learning (CoRL), 2019. [pdf] [appendix]

Adversarial actor-critic method for task and motion planning problems using planning experience.
Beomjoon Kim, Leslie Pack Kaelbling, Tomás Lozano-Pérez
AAAI Conference on Artificial Intelligence (AAAI), 2019. Oral (top 6% of accepted papers). [pdf] [appendix]

Regret bounds for meta Bayesian optimization with an unknown Gaussian process prior.
Zi Wang*, Beomjoon Kim*, Leslie Pack Kaelbling
Neural Information Processing Systems (NeurIPS), 2018. Spotlight (top 3.5% of accepted papers). [pdf]

Guiding search in continuous state-action spaces by learning an action sampler from off-target search experience.
Beomjoon Kim, Leslie Pack Kaelbling, Tomás Lozano-Pérez
AAAI Conference on Artificial Intelligence (AAAI), 2018. Oral (top 6% of accepted papers). [pdf]

Learning to guide task and motion planning using score-space representation.
Beomjoon Kim, Leslie Pack Kaelbling, Tomás Lozano-Pérez
International Conference on Robotics and Automation (ICRA), 2017. Best Cognitive Robotics Paper award. [pdf]

Socially adaptive path planning in human environments using inverse reinforcement learning .
Beomjoon Kim, Joelle Pineau
International Journal of Social Robotics, 2016. [pdf]

Generalizing over uncertain dynamics for on-line trajectory generation.
Beomjoon Kim, Leslie Pack Kaelbling, Tomás Lozano-Pérez
International Symposium of Robotics Research (ISRR), 2015. [pdf]

Learning from limited demonstrations .
Beomjoon Kim, Amir-massoud Farahmand, Doina Precup, Joelle Pineau.
Neural Information Processing Systems (NeurIPS), 2013. Spotlight (top 4% of accepted papers). [pdf]

Maximum mean discrepancy imitation learning.
Beomjoon Kim, Joelle Pineau.
Robotics: Science and Systems (RSS), 2013. [pdf]


template for this site