SDF-Avatar: Animating Generative SDF Grids for One-Shot Avatar
SIGGRAPH Asia 2026
We animate generative SDF grids from a single image, preserving detailed head geometry and stable topology.
Yudong Guo
I am an assistant professor at the University of Science and Technology of China (USTC). My research focuses on 3D vision and digital humans.
I received my Ph.D. from USTC in 2021, advised by Prof. Juyong Zhang, and my bachelor's degree in Statistics from USTC in 2015. From fall 2016 to spring 2017, I was a research assistant in the MultiMedia Lab at Nanyang Technological University, supervised by Prof. Jianfei Cai and Prof. Jianmin Zheng.
Publications
† Corresponding author · * Equal contribution
2026
SplatCtrlA: Generalizable Single Image to Fully Controllable 3D Avatar
ECCV 2026
We create 3D avatars from a single image in seconds, with control over body pose, hand gestures, and facial expressions.
Efficient Large-Scale Scene Reconstruction via Semantic-Aware Hybrid Representation
TVCG 2026
We combine textured meshes and 3D Gaussians through semantic guidance for compact large-scale scene reconstruction and efficient rendering.
ExpPortrait: Expressive Portrait Generation via Personalized Representation
CVPR 2026
We propose a personalized head representation that disentangles expression and identity for highly expressive portrait video synthesis.
Easy3E: Feed-Forward 3D Asset Editing via Rectified Voxel Flow
CVPR 2026
We edit 3D assets in a single feed-forward pass using sparse voxel flow for fast, consistent results.
GSwap: Realistic Head Swapping with Dynamic Neural Gaussian Field
TVCG 2026
We combine a dynamic Gaussian feature field with SMPL-X for consistent video head swapping and natural head-torso blending.
2025
Expressive Talking Human from Single-Image with Imperfect Priors
ICCV 2025
We adopt a 3DGS-Mesh coupled representation with pre-trained human video diffusion models for one-shot vivid talking avatars.
PICA: Physics-Integrated Clothed Avatar
TVCG 2025
We integrate a neural physics simulation engine with the 3DGS avatar for realistic clothed body animation.
2024
Portrait Video Editing Empowered by Multimodal Generative Priors
SIGGRAPH Asia Conference Proceedings 2024
We combine FlashAvatar with multimodal generative models to achieve consistent portrait video editing.
FlashAvatar: High-fidelity Head Avatar with Efficient Gaussian Embedding
CVPR 2024
We embed 3D Gaussians in mesh UV space for head avatar rendering at 300 FPS.
2022
Reconstructing Personalized Semantic Facial NeRF Models From Monocular Video
SIGGRAPH Asia (Journal Track) 2022
2021
Prior-Guided Multi-View 3D Head Reconstruction
TMM 2021
We utilize prior-guided implicit neural rendering to reconstruct the 3D head model.
AD-NeRF: Audio Driven Neural Radiance Fields for Talking Head Synthesis
ICCV 2021
We address the talking head problem with the aid of dynamic neural radiance fields.
3D Face From X: Learning Face Shape from Diverse Sources
TIP 2021
We jointly learn a parametric face model and 3D face reconstruction from diverse sources.
Landmark Detection and 3D Face Reconstruction for Caricature using a Nonlinear Parametric Model
Graphical Models 2021
We automatically detect landmarks and reconstruct 3D caricature faces using a nonlinear parametric model.
2020
Lightweight Photometric Stereo for Facial Details Recovery
CVPR 2020
We present a framework to recover high-fidelity face shapes from one-to-three images captured under near lights.
2018
CNN-based Real-time Dense Face Reconstruction with Inverse-rendered Photo-realistic Face Images
TPAMI 2018
We train 3DFaceNet with generated face images for real-time dense reconstruction from monocular RGB video.