Seong Hyeon Park

Note: I am currently on the job market.

Visiting Scholar at the Global AI Frontier Lab, New York University (NYU).

My research focuses on efficient algorithms for robot learning and high-dimensional computer vision. In particular, I have worked on dexterous (high-degree-of-freedom) robotic hand policies, 3D/4D reconstruction, and video motion estimation.

I earned my Ph.D. at KAIST under the supervision of Dr. Jinwoo Shin, where I also collaborated closely with Dr. Paul Hongsuck Seo and Dr. Sukmin Yun. Prior to my Ph.D., I developed vehicle and pedestrian motion prediction algorithms under the academic supervision of Dr. Jun-Won Choi.

Publications (P: preprints; C: conferences; W: workshops)

[P1] Dexterous Point Policy: Learning Point-based Dexterous Hand Policies from Human Demonstrations (2026)

Beomjun Kim*, Seong Hyeon Park*, Seunghoon Sim, SeungJun Moon, SangHyeok Lee and Jinwoo Shin (∗ Contributed equally)

[P2] Pose6DAug: Physically Plausible Multi-View Object Swapping for Robot Data Augmentation (2026)

Jonghoon Lee, Seong Hyeon Park, Minha Lee, Byungwoo Jeon and Jinwoo Shin

[C8] Track3R: Joint Point Map and Trajectory Prior for Spatiotemporal 3D Understanding

Seong Hyeon Park and Jinwoo Shin

NeurIPS 2025

[C7] TrackIME: Enhanced Video Point Tracking via Instance Motion Estimation

Seong Hyeon Park, Huiwon Jang, Byungwoo Jeon, Sukmin Yun, Paul Hongsuck Seo and Jinwoo Shin

NeurIPS 2024 (spotlight)

[C6] Tabular Transfer Learning via Prompting LLMs

Jaehyun Nam, Woomin Song, Seong Hyeon Park, Jihoon Tack, Sukmin Yun, Jaehyung Kim, Kyu Hwan Oh and Jinwoo Shin

Conference on Language Modeling (COLM) 2024

[C5] IFSeg: Image-free Semantic Segmentation via Vision-Language Model

Sukmin Yun*, Seong Hyeon Park*, Paul Hongsuck Seo and Jinwoo Shin (∗ Contributed equally)

CVPR 2023

[C4] K-centered Patch Sampling for Efficient Video Recognition

Seong Hyeon Park, Jihoon Tack, Byeongho Heo, Jung-Woo Ha and Jinwoo Shin

ECCV 2022

[C3] LaPred: Lane-Aware Prediction of Multi-Modal Future Trajectories of Dynamic Agents

ByeongDo Kim, Seong Hyeon Park, Seokhwan Lee, Elbek Khoshimjonov, Dongsuk Kum, Junsoo Kim, Jeong Soo Kim and Jun-Won Choi

CVPR 2021

[C2] Diverse and Admissible Trajectory Forecasting through Multimodal Context Understanding

Seong Hyeon Park, Gyubok Lee, Minseok Kang, Jimin Seo, Manoj Bhat, Ashwin Jadhav, Jonathan Francis, Paul Pu Liang and Louis-Philippe Morency

ECCV 2020

[C1] Sequence-to-Sequence Prediction of Vehicle Trajectory via LSTM Encoder-Decoder Architecture

Seong Hyeon Park, Byeongdo Kim, Chang Mook Kang, Chung Choo Chung and Jun-Won Choi

IEEE Intelligent Vehicles Symposium (IV) 2018 (oral presentation)

[W1] MC2: Multi-view Consistent Depth Estimation via Coordinated Image-based Neural Rendering

Subin Kim*, Seong Hyeon Park*, Sihyun Yu, Kihyuk Sohn and Jinwoo Shin (∗ Contributed equally)

CVPR Workshop on Neural Rendering Intelligence 2024