Postdoctoral Researcher, PI Lab, Seoul National University

Jongmin Park

I reconstruct dynamic 3D scenes from ordinary videos, and I'm now using those representations to help embodied agents understand and act in the world.

Portrait of Jongmin Park

About

I am a postdoctoral researcher at PI Lab, Seoul National University, working with Prof. Youngjae Yu on world action models and vision-language-action models for embodied agents.

I received my Ph.D. in Electrical Engineering from KAIST in 2026, advised by Prof. Munchurl Kim at VICLAB. My research centers on 3D scene reconstruction with neural radiance fields and 3D Gaussian splatting: real-time dynamic 3D Gaussians from monocular video, robust reconstruction from blurry footage, and efficiency-controllable feed-forward 3D Gaussian splatting from multi-view images.

Work experience

  • 2026–PresentPostdoc, PI Lab,
    Seoul National University

Education

  • 2022–2026Ph.D., Electrical Engineering, KAIST
  • 2020–2022M.S., Electrical Engineering, KAIST
  • 2015–2020B.S., Electrical Engineering, KAIST

Published at

CVPRICCVAAAIIEEE TPAMI

Research interests

Dynamic 3D reconstructionFeed-forward 3D Gaussian splatting World action modelsVision-language-action models

Publications

* equal contribution

PreprintarXiv, June 2026 · under review

WebSpline: Structure-Informed Splines for Real-Time 3D Gaussians from Monocular Videos

Jongmin Park*, Jeonghwan Yun*, Minh-Quan Viet Bui, Munchurl Kim

CVPR 2026Highlighttop 3.6% of 16,092 submissions

EcoSplat: Efficiency-controllable Feed-forward 3D Gaussian Splatting from Multi-view Images

Jongmin Park*, Minh-Quan Viet Bui*, Juan Luis Gonzalez Bello, Jaeho Moon, Jihyong Oh, Munchurl Kim

AAAI 2026acceptance rate 17.6%

MoBGS: Motion Deblurring Dynamic 3D Gaussian Splatting for Blurry Monocular Video

Minh-Quan Viet Bui*, Jongmin Park*, Juan Luis Gonzalez Bello, Jaeho Moon, Jihyong Oh, Munchurl Kim

CVPR 2025acceptance rate 22.1%

SplineGS: Robust Motion-Adaptive Spline for Real-Time Dynamic 3D Gaussians from Monocular Video

Jongmin Park*, Minh-Quan Viet Bui*, Juan Luis Gonzalez Bello, Jaeho Moon, Jihyong Oh, Munchurl Kim

IEEE TPAMI 2025vol. 47, no. 9, pp. 7752–7770

MoBluRF: Motion Deblurring Neural Radiance Fields for Blurry Monocular Video

Minh-Quan Viet Bui*, Jongmin Park*, Jihyong Oh, Munchurl Kim

ICCV 2023acceptance rate 25.0%

COMPASS: High-Efficiency Deep Image Compression with Arbitrary-scale Spatial Scalability

Jongmin Park, Jooyoung Lee, Munchurl Kim

Workshop and domestic publications
  • C. O. Ancuti, C. Ancuti, F.-A. Vasluianu, R. Timofte, and 48 others including J. Park. NTIRE 2020 Challenge on NonHomogeneous Dehazing. CVPR Workshops, 2020.
  • J. Park, M. Kim. Transformer-based Image Compression Framework of Spatial Scalability. KIBME Workshops, 2023.
  • J. Park, G. Youk, M. Kim. SAR Image Target Detection based on Attention YOLOv4. Journal of KIMST, vol. 25, no. 5, 2022.

Patents & awards

Patents

  • 2025Curve Model Based Rendering Method and Apparatus for Performing the Same KR10-2025-0189396, filed
  • 2024Method, Apparatus and Recording Medium for Encoding/Decoding Image KR10-2024-0071361, filed
  • 2023Method and Apparatus for Image Translation Using Latent Features of Image KR10-2563953, registered

Awards & scholarships

  • 2026CVPR Highlight paper (EcoSplat)
  • 2023Best Paper Award (Chang-kyu Park Academic Award) Korea Institute of Military Science and Technology
  • 2023Outstanding Paper Award Korea Institute of Broadcast and Media Engineers
  • 2020–26Government Scholarship for M.S. and Ph.D. studies
  • 2015–20National Scholarship for B.S. studies

Curriculum vitae

Education, publications, research projects, and patents in one PDF. Last updated September 2026.

Contact

I'm happy to talk about collaborations on dynamic 3D reconstruction, feed-forward Gaussian splatting, or embodied AI.

jm.park@snu.ac.kr

Elsewhere