About Myself
I am Jongmin Park, a Ph.D. student in the Department of Electrical Engineering at the Korea Advanced Institute of Science and Technology (KAIST), advised by Prof. Munchurl Kim at VICLAB.
My research interests lie in 3D computer vision, novel view synthesis (NeRF, Gaussian Splatting), low-level vision, and generative models.
Email: jm.park@kaist.ac.kr
[CV] [Google Scholar] [LinkedIn] [Github]
Education
09.2022 - Present Ph.D. candidate in School of Electrical Engineering, KAIST
Advisor: Prof. Munchurl Kim (VICLAB)
09.2020 - 08.2022 M.S. in School of Electrical Engineering, KAIST
Advisor: Prof. Munchurl Kim (VICLAB)
03.2015 - 08. 2020 B.S. in School of Electrical Engineering, KAIST
Preprints
🕵🏻♂️ 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☨ and Munchurl Kim☨
(*: equal contribution, ☨: co-corresponding author)
(under review) arXiv 2025.
[arxiv] [Project Page]
Publications
MoBluRF: Motion Deblurring Neural Radiance Fields for Blurry Monocular Video
Minh-Quan Viet Bui*, Jongmin Park*, Jihyong Oh☨ and Munchurl Kim☨
(*: equal contribution, ☨: co-corresponding author)
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2025. [AI Top-tier Journal (IF=20.8)]
Accepted for publication
[arxiv] [Project Page]
∿ 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☨ and Munchurl Kim☨
(*: equal contribution, ☨: co-corresponding author)
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025. [AI Top-tier Conf.]
(Acceptance Rate: 22.1% ≈ 2,878/13,008)
[arxiv] [Project Page] [Code]
COMPASS: High-Efficiency Deep Image Compression with Arbitrary-scale Spatial Scalability
Jongmin Park, Jooyoung Lee and Munchurl Kim
IEEE/CVF Conference on Computer Vision and Pattern Recognition (ICCV), 2023. [AI Top-tier Conf.]
(Acceptance Rate: 25.0% ≈ 2,155/8,620)
[Paper] [Project Page] [Code]
Research Projects
04.2022 - Present Deep Learning-based Rendering of Spatial Video with Stationary and Dynamic Scenes supported by SW STAR LAB
Proposed a COLMAP-free dynamic 3D reconstruction framework via motion-adaptive spline deformation of 3D Gaussians.
Investigated and developed novel methods for dynamic 3D reconstruction from blurry monocular videos, comparing different neural representations.
Initially proposed a NeRF-based framework to achieve high-fidelity novel view synthesis, establishing a strong baseline for the deblurring task.
Subsequently developed a 3D Gaussian Splatting-based method, which enabled superior sharpness and rendering speed over the NeRF approach.
09.2022 – 05.2023 AI-based Image Compression with Spatial Scalability supported by Electronics and Telecommunications Research Institute
Proposed a learned, scalable image compression framework capable of decoding at any arbitrary (non-integer) scale from a single bitstream.
Patents
KR1025639530000: Method and Apparatus for Image Translation Using Latent Features of Image, 08. 2023
KR1020240071361: METHOD, APPARATUS AND RECORDING MEDIUM FOR ENCODING/DECODING IMAGE ,05. 2024
Honors & Awards
Scholarships
09.2022 – Present Government Scholarship for Ph.D. studies
09.2020 – 08.2022 Government Scholarship for M.S. studies
03.2015 – 08.2020 National Scholarship for B.S. studies
Awards
06.2023 Best Paper Award (Chang-kyu Park Academic Award) at Korea Institute of Military Science and Technology
06.2023 Outstanding Paper Award at Korea Institute of Broadcast and Media Engineers