Postdoctoral Fellow
AI Institute, Seoul National University
chonkang at snu dot ac dot kr

Bio

I recently received my PhD in AI from Seoul National University, advised by Prof. Byoung-Tak Zhang. My research lies at the intersection of robot learning and multimodal AI, where I focus on building generalist robots by leveraging semantic priors from large pretrained models. Specific topics include:

• Robotics Foundation Models: Building vision-language-action (VLA) models for generalist robot policies, with a focus on grounding web-scale knowledge to robotic control (CLIP-RT).
• Embodied Reasoning: Investigating vision-language models (VLMs) for task planning (Socratic Planner) and reasoning about language instructions (PROGrasp, PGA, GVCCI).
• Vision & Language: Developing interactive VLMs that engage in continuous, grounded communication with humans about images (GST, SGL, DAN) and videos (MASN).

My PhD research has been supported by fellowships from Youlchon Foundation and IPAI. I was fortunate to collaborate with researchers in NAVER AI and SK T-Brain.

Prior to joining PhD, I did my master study in Cognitive Science at Seoul National University. Studying cognitive science has sparked my interest in AI and interdisciplinary research. I earned my Bachelor’s degree in Computer Science from Ajou University.

I'm on the job market! Please let me know if you have any opportunities for me.

News

• [April 2025] Happy to announce that our work (CLIP-RT) is accepted to RSS 2025!
• [April 2025] I’m selected as a member of RSS Pioneers!
• [Nov 2024] Happy to release a new robotics foundation model, CLIP-RT!
• [Aug 2024] I’m selected as a recipient of the Youlchon AI Star Fellowship.
• [Jun 2024] PGA is accepted at IROS 2024.
• [Apr 2024] A preprint for embodied instruction following (Socratic Planner) is released.
• [Mar 2024] I wrote my research statement about what I’ve been studying.
• [Mar 2024] A new preprint (Continual Vision-and-Language Navigation) is released.
• [Jan 2024] PROGrasp is accepted to ICRA 2024!
• [Dec 2023] I attend Brainlink 2023.
• [Nov 2023] I’ll give a talk at Dept. of Energy Resources Engineering at Seoul National University (Title: “The Evolution of Language Models: From Basic NLP to ChatGPT and Beyond”).
• [Oct 2023] Two preprints (PROGrasp and PGA) are released!
• [Jun 2023] One paper is accepted to IROS 2023!
• [Mar 2023] Happy to announce that our paper is accepted to CVPR 2023!
• [Jun 2022] One paper is accepted to ICML 2022 Pre-training Workshop.
• [May 2022] Thrilled to announce that our new preprint is released!
• [Apr 2022] One paper is accepted to CVPR 2022 HCIS Workshop.
• [Dec 2021] I gave an invited talk at Korea Software Congress.
• [Oct 2021] One paper is accepted to NeurIPS 2021 CtrlGen Workshop.
• [Aug 2021] One paper is accepted to Findings of EMNLP 2021.
• [May 2021] One paper is accepted to ACL 2021.
• [Sep 2020] I’m starting my Ph.D. in this fall.
• [Jun 2020] From July, I’ll join SNU AI Institute (AIIS) as a researcher.
• [Jan 2020] Our paper has been accepted to ICASSP 2020!
• [Dec 2019] From January, I’ll be a research intern at SK T-Brain!
• [Nov 2019] I gave a spotlight talk at Video Turing Test workshop, ICCV 2019.
• [Oct 2019] I gave an invited talk at SK Telecom AI Center.
• [Aug 2019] Excited to announce that our paper has been accepted to EMNLP 2019.
• [Jun 2019] Our proposed method ranks 3rd place in Visual Dialog Challenge 2019!!
• [Aug 2018] We have a paper accepted to ECCV 2018 Workshop on VizWiz Grand Challenge.

Highlights (Here are all papers)

CLIP-RT: Learning Language-Conditioned Robotic Policies from Natural Language Supervision

Gi-Cheon Kang^*, Junghyun Kim^*, Kyuhwan Shim, Jun Ki Lee^†, Byoung-Tak Zhang^†
RSS 2025
CoRL 2024 Workshop on Language and Robot Learning Project Page Paper Code

PROGrasp: Pragmatic Human-Robot Communication for Object Grasping

Gi-Cheon Kang, Junghyun Kim, Jaein Kim, Byoung-Tak Zhang
ICRA 2024 (Oral)
Paper Code

The Dialog Must Go On: Improving Visual Dialog via Generative Self-Training

Gi-Cheon Kang, Sungdong Kim^*, Jin-Hwa Kim^*, Donghyun Kwak^*, Byoung-Tak Zhang
CVPR 2023
ICML 2022 Workshop on Pre-training: Perspectives, Pitfalls, and Paths Forward Project Page Paper Code Slides Video

Invited Talks

Services