Min-Seop Kwak

Email: minseop [dot] kwak [at] kaist [dot] ac [dot] kr
Google Scholar / Github / Twitter / LinkedIn / CV

I am a Ph.D. student at KAIST AI under the supervision of Prof. Seungryong Kim. I am currently a visiting scholar at the University of Michigan, supervised by Prof. Jeong Joon Park. Prior to that, I conducted a research internship at NAVER AI Lab under the supervision of Dr. Jin-Hwa Kim, collaborating with Dr. Sangdoo Yun and Dr. Dongyoon Han on MoAI, a diffusion-based novel view synthesis project.

My current main research interest is controllable and consistent 3D / video generation, to enhance the usability and realism of generated scenes. I also have extensive experience with all major forms of neural 3D representations, especially 3DGS and NeRF. If you have anything to ask, such as collaboration, please contact me freely!

"All this is from God, who reconciled us to himself through Christ and gave us the ministry of reconciliation."
2 Corinthians 5:18

Publications ( show selected / show by date )

(* indicates equal contribution and ^† indicates equal advising)

🗿 MoAI: Aligned Novel View Image and Geometry Synthesis via Cross-modal Attention Instillation

Min-Seop Kwak, Junho Kim, Sangdoo Yun, Dongyoon Han, Taekyung Kim, Seungryong Kim^†, Jin-Hwa Kim^†

ICLR 2026 / Website / Paper / Code

Self-Evolving Neural Radiance Fields

Jaewoo Jung, Jisang Han, Jiwon Kang*, Seongchan Kim, Min-Seop Kwak, Seungryong Kim

Wild3D Workshop @ ICCV 2025 / Website / Paper / Code

Geometry-Aware Score Distillation via 3D Consistent Noising and Gradient Consistency Modeling

Min-Seop Kwak, Donghoon Ahn, Hyeonsu Kim, Jin-Hwa Kim^†, Seungryong Kim^†

ArXiv Preprint 2025 / Website / Paper / Code

Retrieval-Augmented Score Distillation for Text-to-3D Generation

Junyoung Seo, Susung Hong, Wooseok Jang*, Hyeonsu Kim, Min-Seop Kwak, Doyup Lee^†, Seungryong Kim^†

ICML 2024 / Website / Paper / Code

3DFuse: Let 2D Diffusion Model Know 3D Consistency for Robust Text-to-3D Generation

Junyoung Seo, Wooseok Jang, Min-Seop Kwak*, Hyeonsu Kim, Jaehoon Ko, Junho Kim, Jin-hwa Kim^†, Jiyoung Lee^†, Seungryong Kim^†

ICLR 2024 / Website / Paper / Code

DäRF: Boosting Radiance Fields from Sparse Inputs with Monocular Depth Adaptation

Jiuhn Song, Seonghoon Park, Honggyu An*, Seokju Cho, Min-Seop Kwak, Sungjin Cho, Seungryong Kim

NeurIPS 2023 / Website / Paper / Code

GeCoNeRF: Few-shot Neural Radiance Fields via Geometric Consistency

Min-Seop Kwak, Jiuhn Song, Seungryong Kim

ICML 2023 / Website / Paper / Code

Experience

Visiting Scholar, University of Michigan

Aug 2025 - Aug. 2026

Working on image/video generative model and robot learning, advised by Prof. Jeong Joon Park.

Research Intern, Naver AI Lab

Jul. 2024 - Jan. 2025

Worked on MoAI, a generative image+geometry NVS model, under the supervision of Jin-Hwa Kim.

Teaching

Teaching Assistant, Samsung AI Expert Program

Jul. 2025, Jul. 2024, Aug. 2023

Transformer-based Image Classification, Fall 2025
Image classification, Fall 2024
Towards Better Conditional (or Guided) Generative Models, Fall 2023

Education

KAIST, Seoul, Korea

Ph.D. In Artificial Intelligence (Aug. 2024 - Present)

Korea University, Seoul, Korea

B.S. in Mechanical Engineering (Mar. 2015 - Feb. 2022)
M.S./Ph.D. Integrated Program in Computer Science and Engineering
(Mar. 2022 - Aug. 2024, incomplete)