Research Area
-
Research Overview

-
Relightable and Dynamic 3DGS Human Avatar

- Monocular video-based Human Avatar Reconstruction
- 3DGS-based Dynamic Avatar Representation
- Novel Pose Synthesis- Photorealistic Relighting under varying illumination conditions
-
Multi-view Camera Studio

- Dome-type Light Stage Capture Studio
- High resolution 3D Face & Body Capture
- Lighting Control System
- Integrated Audio-Visual Capture Environment
-
3D Animatable Head Avatar

- Monocular Video–based Head Avatar Modeling
- Disentangled Static, Dynamic, Geometric Features
- Cross-identity Facial Motion Transfer
- High-fidelity Facial Expression Animation
-
3D Motion Capture Assessment

- High-Fidelity Ground-Truth Motion Capture and Data Acquisition
- SMPL-based Full-body Reconstruction and Biomechanical Modeling
- Music-Conditioned Dance Enhancement and Feature Analysis
- LLM-based Interpretation and Qualitative Evaluation
