Jiye Lee

Hi, I'm Jiye Lee. I'm a PhD candidate at Seoul National University. I am a member of SNU Intelligent Motion Lab, advised by Prof. Jungdam Won.

Previously, I was advised by Prof. Hanbyul Joo in the early stage of my PhD. I was also fortunate to visit Meta Codec Avatars Lab as a Research Scientist Intern in 2024.

     
RESEARCH

I focus on understanding and reconstructing human movements from various levels of world perceptions. This involves interpreting human motions from visual or wearable sensors (sensory perception) and synthesizing realistic and diverse human-scene interactions (virtual perception of the environment).

UPDATES
Sep 2025
One first-author paper is accepted to SIGGRAPH Asia 2025.
Jun 2025
I am co-organizing a workshop on Global 3D Human Poses in conjunction with CVPR 2025.
Feb 2024
One first-author paper is accepted to CVPR 2024.
PUBLICATIONS

Audio Driven Real-Time Facial Animation for Social Telepresence

Jiye Lee, Chenghui Li, Linh Tran, Shih-En Wei, Jason Saragih, Alexander Richard, Hanbyul Joo and Shaojie Bai

SIGGRAPH Asia 2025  /   Project   Paper

Real-time audio-driven 3D facial animation using online transformers and single-step diffusion distillation to minimize latency. Work done with Meta Codec Avatars Lab.

Learning to Transfer Human Hand Skills for Robot Manipulations

Sungjae Park*, Seungho Lee*, Mingi Choi*, Jiye Lee, Jeonghwan Kim, Jisoo Kim and Hanbyul Joo

CoRL X-Embodiment Workshop 2024  /   Project   Paper

Dexterous robot manipulation retargeted from human hand mocap via joint human-robot-object motion manifolds.

Mocap Everyone Everywhere: Lightweight Motion Capture With Smartwatches and a Head-Mounted Camera

Jiye Lee and Hanbyul Joo

CVPR 2024  /   Project   Paper   Code

Lightweight full-body motion capture with two smartwatches and a head-mounted camera, applicable to diverse indoor and outdoor scenarios.

Locomotion-Action-Manipulation: Synthesizing Human-Scene Interactions in Complex 3D Environments

Jiye Lee and Hanbyul Joo

ICCV 2023  /   Project   Paper   Code

Human motion synthesis in complex indoor environments without using paired scene-motion data via RL-based test-time optimization.

Learning Virtual Chimeras by Dynamic Motion Reassembly

Seyoung Lee, Jiye Lee and Jehee Lee

ACM ToG (SIGGRAPH Asia 2022)  /   Project   Paper   Code

Animating chimeras by reassembling body parts and movements of source characters via a part assembler and an RL-based dynamic controller.

ACADEMIC SERVICES

Workshop Organizer


Journal Reviewer

  • ACM Transactions on Graphics (SIGGRAPH Asia) 2024, 2025
  • Transactions on Visualization and Computer Graphics (TVCG) 2025

Conference Reviewer

  • CVPR, ICCV, ICRA, NeurIPS, 3DV, BMVC, WACV