Integrated M.S./Ph.D. Student · VDSLab, Sogang University

Suwoong Yeom

I received my B.S. in Mechanical Engineering from Ajou University, and I am now an integrated M.S./Ph.D. student in Electronic Engineering at Sogang University, advised by Prof. Suk-Ju Kang. I also work with Prof. Kyeongbo Kong on collaborative research.

My research has centered on reconstructing 3D/4D space with NeRF and Gaussian Splatting. I currently work on spatial AI, studying how feed-forward models come to understand spatial structure, and on world models for digital twins that supply training data and simulation for physical AI.

  • 3D/4D Reconstruction
  • 3D/4D Computer Vision
  • Spatial AI
  • World Model
Email Google Scholar GitHub LinkedIn CV (PDF)

Education

Degrees only. Visiting positions live under Experience

2023.03 —

Integrated M.S./Ph.D. in Electronic Engineering

Sogang University · VDSLab

AdvisorProf. Suk-Ju Kang

2016.03 — 2022.08

B.S. in Mechanical Engineering

Ajou University

Experience

Visiting positions and internships

2026.06 — 2026.07

Visiting Researcher

Purdue University · NEISLab

HostProf. Younghyun Kim

Selected Work

Projects I led or co-led

ECCV 2026

TRiGS: Temporal Rigid-Body Motion for Scalable 4D Gaussian Splatting

Rigid-body motion priors that keep 4D Gaussian splatting tractable as scenes grow.

arXiv Project Code

IEEE TMM · Submitted

HOIGS: Human-Object Interaction Gaussian Splatting

Joint reconstruction of humans and the objects they interact with, in one splatting model.

arXiv

IEEE TCSVT · Submitted

EMDGS: Enhancing Monocular Dynamic Gaussian Splatting via Adaptive Pipeline Integration

Monocular dynamic reconstruction that needs no SfM poses or point clouds. Static and dynamic regions are separated explicitly, and the whole pipeline drops into existing 4DGS methods.

Publications

* equal contribution  ·  † co-corresponding author

Conference2

C2

European Conference on Computer Vision (ECCV 2026)

TRiGS: Temporal Rigid-Body Motion for Scalable 4D Gaussian Splatting

*Suwoong Yeom, *Joonsik Nam, *Seunggyu Choi, Lucas Yunkyu Lee, Sangmin Kim, Jaesik Park, Joonsoo Kim, Kug-jin Yun, Kyeongbo Kong†, Suk-Ju Kang†

arXiv Project Page Code

C1

Motion mask pipeline figure

International Conference on Electronics, Information and Communications (ICEIC 2024)

Motion Mask-driven Improvements in Monocular Dynamic Novel View Synthesis

Suwoong Yeom, Hosung Son, Chanhee Kang, Joonsoo Kim, Kug-jin Yun, Won-Sik Cheong, Suk-Ju Kang

IEEE Xplore

Journal3

J3

ICSD-NeRF pipeline figure

IEEE Transactions on Computational Imaging (TCI)

ICSD-NeRF: Independent Canonical Spaces for Enhanced Dynamic Scene Modeling in Neural Radiance Fields

Suwoong Yeom, Hosung Son, Chanhee Kang, Eunho Shin, Joonsoo Kim, Kug-jin Yun, Suk-Ju Kang

IEEE Xplore Code

J2

IEEE Transactions on Circuits and Systems for Video Technology (TCSVT) Under Review

EMDGS: Enhancing Monocular Dynamic Gaussian Splatting via Adaptive Pipeline Integration

*Suwoong Yeom, *Jimin Roh, *Eunho Shin, Songju Na, Joonsoo Kim, Kug-jin Yun, Kyeongbo Kong†, Suk-Ju Kang†

J1

IEEE Transactions on Multimedia (TMM) Under Review

HOIGS: Human-Object Interaction Gaussian Splatting

*Taewoo Kim, *Suwoong Yeom, *Jaehyun Pyun, Geonho Cha, Dongyoon Wee, Joonsik Nam, Yun-Seong Jeong, Kyeongbo Kong†, Suk-Ju Kang†

arXiv

Mentoring

Junior researchers I worked closely with, and the papers that came out of it

2026

Attention-gated capacity allocation pipeline figure

MenteeSeunggyu ChoiM.S. Student

European Conference on Computer Vision (ECCV 2026) 3DWM Workshop

The Decoder Already Knows Where to Densify: Attention-Gated Capacity Allocation for Feed-Forward 3D Gaussian Splatting

*Seunggyu Choi, *Suwoong Yeom, Suk-Ju Kang†

PDF Workshop Page

2026

MenteeChaewon LeeUndergraduate Researcher

European Conference on Computer Vision (ECCV 2026) TwinWorld Workshop

Seamless Spatial Extension in 3D Gaussian Splatting via Auxiliary Wild Views

*Chaewon Lee, *Seunggyu Choi, *Suwoong Yeom, Joonsoo Kim, Kugjin Yun, Suk-Ju Kang†

PDF OpenReview

2026

Dual-PAG pipeline figure

MenteeJeongseon OhIntegrated M.S./Ph.D. Student

International Conference on Machine Learning (ICML 2026) F2S Workshop

Dual-PAG: Bidirectional Perturbed-Attention Guidance for Residue-Free Video Editing

*Jeongseon Oh, *Suwoong Yeom, *ChangHee Yang, Suk-Ju Kang†

PDF

Research Projects

Funded and industry collaborations

June 2025 —

NAVER CLOUD 공동 연구

NAVER Cloud, Korea · 참여연구원

Feed-Forward 3D/4D Reconstruction World Modeling

Mar 2023 —

초실감 메타버스 서비스를 위한 실사기반 입체영상 공간컴퓨팅 기술 개발

ETRI, Korea · 참여연구원

3D/4D Reconstruction Depth Estimation Neural Rendering