Junwoo Chang

I am an incoming Ph.D. student in Mechanical Engineering at UC Berkeley, where I will be advised by Prof. Negar Mehr. I obtained my M.S. and B.S. at Yonsei University under the guidance of Prof. Jongeun Choi. My research focuses on robotics and reinforcement learning, especially on developing sample-efficient, generalizable, and theoretically grounded methods.

Email / CV / Scholar / Github

News

Fall 2026: I will join UC Berkeley as a Ph.D. student in Mechanical Engineering.
Apr. 2026: Our paper, "Symmetry-Aware Steering of Equivariant Diffusion Policies: Benefits and Limits", was accepted to IFAC 2026.
Apr. 2026: Our paper, "Multi-Robot Motion Planning from Vision and Language using Heat-Inspired Diffusion", was accepted to IEEE RA-L.
Jan. 2026: Our paper, "Partially Equivariant Reinforcement Learning in Symmetry-Breaking Environments", was accepted to ICLR 2026.
Mar. 2025: Our tutorial paper on geometric deep learning, "SE(3)-Equivariant Robot Learning and Control: A Tutorial Survey", was published in IJCAS.
Apr. 2024: Our paper, "Diffusion-EDFs: Bi-equivariant Denoising Generative Modeling on SE(3) for Visual Robotic Manipulation", was selected for a Highlight at CVPR 2024.
Oct. 2023: Our paper, "Denoising Heat-inspired Diffusion with Insulators for Collision-Free Motion Planning", was accepted to the NeurIPS 2023 Workshop on Diffusion Models.

Publications

( ^* Equal contribution, ^† Equal advising)

	Symmetry-Aware Steering of Equivariant Diffusion Policies: Benefits and Limits Minwoo Park^, Junwoo Chang^, Jongeun Choi, Roberto Horowitz IFAC 2026 paper \| code
	Multi-Robot Motion Planning from Vision and Language using Heat-Inspired Diffusion Jebeom Chae^, Junwoo Chang^, Seungho Yeom, Yujin Kim, Jongeun Choi IEEE RA-L, accepted, 2026 paper
	Partially Equivariant Reinforcement Learning in Symmetry-Breaking Environments Junwoo Chang, Minwoo Park, Joohwan Seo, Roberto Horowitz, Jongmin Lee^†, Jongeun Choi^† ICLR 2026 project page \| paper \| code (TBD)
	Group-Invariant Unsupervised Skill Discovery: Symmetry-aware Skill Representations for Generalizable Behavior Junwoo Chang, Joseph Park, Roberto Horowitz, Jongmin Lee^†, Jongeun Choi^† Preprint (Under Review), 2026 paper \| code (TBD)
	SE(3)-Equivariant Robot Learning and Control: A Tutorial Survey Joohwan Seo, Soochul Yoo, Junwoo Chang, Hyunseok An, Hyunwoo Ryu, Soomi Lee, Arvind Kruthiventy, Jongeun Choi, Roberto Horowitz, IJCAS 2025 paper
	Diffusion-EDFs: Bi-equivariant Denoising Generative Modeling on SE(3) for Visual Robotic Manipulation Hyunwoo Ryu, Jiwoo Kim, Hyunseok An, Junwoo Chang, Joohwan Seo, Taehan Kim, Yubin Kim, Chaewon Hwang, Jongeun Choi, Roberto Horowitz, CVPR 2024 (Highlight) project page \| paper \| code
	Denoising Heat-inspired Diffusion with Insulators for Collision Free Motion Planning Junwoo Chang^, Hyunwoo Ryu^, Jiwoo Kim, Joohwan Seo, Nikhil Potu Surya Prakash, Jongeun Choi, Roberto Horowitz, NeurIPS 2023 Workshop on Diffusion Models project page \| paper \| code
Website source code credit to Jon Barron.