Hyunwook Choi

I'm an M.S./Ph.D. student in Artificial Intelligence at KAIST AI, advised by Professor Seungryong Kim. My research lies in computer vision, with a focus on World Models, Auto-Regressive Video Generation, and Efficient Video Generation.

My current research focuses on building video world models. I’m particularly interested in real-world grounded world modeling (Seoul World Model), persistent world modeling (World Observer), embodied and multi-agent world modeling (Multi-Agent Egocentric World Model), and efficient autoregressive generation (Speculative Forcing).

Email / CV / Scholar / LinkedIn / Twitter

News

Oct 2026. World Observer, Speculative Forcing, and ME-World are released on arXiv.
Jul 2026. Seoul World Model is selected for an Oral presentation at ECCV 2026.
Jul 2026. I'm excited to join NAVER AI Lab as a Research Scientist Intern.
Jul 2026. I will present Seoul World Model in the Expo panel talk at ICML 2026.
Jun 2026. Seoul World Model was highlighted in the NVIDIA Newsroom.
Jun 2026. Seoul World Model is selected for an Oral at the F2S Workshop @ ICML 2026.
Mar 2026. I started my M.S./Ph.D. in AI at KAIST AI, advised by Prof. Seungryong Kim.
Jan 2026. MATRIX is accepted to ICLR 2026.
Jun 2025. S4M is accepted to ICCV 2025.

Work Experience

NAVER, Research Scientist Intern
Jul 2026 - Nov 2026
NAVER AI Lab Mentors: Sangdoo Yun, Dongyoon Han

Education

Korea Advanced Institute of Science and Technology (KAIST)
M.S./Ph.D. Integrated Program in Artificial Intelligence · Advisor: Seungryong Kim
Mar 2026 – Present

Korea University
B.S. in Cyber Defence · GPA: 4.33/4.5 (Ranked 1st)
Mar 2022 – Feb 2026
Full-funded Scholarship

Research Experience

Undergraduate Research Intern · Kim Jaechul Graduate School of AI, KAIST
Advisor: Seungryong Kim
Jul 2024 – Feb 2026

Publications

World Observer: Joint Actor-Observer Generation for Persistent World Modeling
	World Observer: Joint Actor-Observer Generation for Persistent World Modeling Hyunwook Choi, Dahyun Chung, Hyunsung Kim, Siyoon Jin, Jinhyeok Choi, Junyoung Seo, Seungryong Kim arXiv preprint, 2026 project page / paper
Speculative Forcing for Efficient Autoregressive Video Generation
	Speculative Forcing for Efficient Autoregressive Video Generation Hyunwook Choi, Seungryong Kim, Dongyoon Han, Sangdoo Yun arXiv preprint, 2026 project page / paper
ME-World: Multi-Agent Egocentric World Model with Fine-Grained Embodied Interaction
	ME-World: Multi-Agent Egocentric World Model with Fine-Grained Embodied Interaction Dahyun Chung, Siyoon Jin, Hyunwook Choi, Honggyu An, Junyoung Seo, Hyunsung Kim, Seung Wook Kim, Seungryong Kim arXiv preprint, 2026 project page / paper
Grounding World Simulation Models in a Real-World Metropolis
	Grounding World Simulation Models in a Real-World Metropolis Junyoung Seo, Hyunwook Choi, Minkyung Kwon, Jinhyeok Choi, Siyoon Jin, Gayoung Lee, Junho Kim, JoungBin Lee, Geonmo Gu, Dongyoon Han, Sangdoo Yun, Seungryong Kim, Jin-Hwa Kim ECCV, 2026 (Oral) / F2S Workshop @ ICML, 2026 (Oral) project page / paper
MATRIX: Mask Track Alignment for Interaction-Aware Video Generation
	MATRIX: Mask Track Alignment for Interaction-Aware Video Generation Siyoon Jin, Seongchan Kim, Dahyun Chung, Jaeho Lee, Hyunwook Choi, Jisu Nam, Jiyoung Kim, Seungryong Kim ICLR, 2026 project page / paper
S4M: Boosting Semi-Supervised Instance Segmentation with Segment Anything Model
	S4M: Boosting Semi-Supervised Instance Segmentation with Segment Anything Model Heeji Yoon, Heeseong Shin, Eunbeen Hong, Hyunwook Choi, Hansang Cho, Daun Jeong, Seungryong Kim ICCV, 2025 project page / paper
Referring Video Object Segmentation via Language-aligned Track Selection
	Referring Video Object Segmentation via Language-aligned Track Selection Seongchan Kim, Woojeong Jin, Sangbeom Lim, Heeji Yoon, Hyunwook Choi, Seungryong Kim AdaptFM Workshop @ ICML, 2026 project page / paper

Website template from Jon Barron.