Xu Chen

I am a Research Scientist in the AR Perception group at Google. I am interested in 3D digital humans, including their animation, reconstruction and generation. I have been working on neural implicit representations and generative models to enable high-fidelity 3D human reconstruction and animation.

I received my PhD degree from the CLS program with ETH Zurich and Max Planck Institute for Intelligent Systems, where I was advised by Prof. Otmar Hilliges and co-supervised by Prof. Michael J. Black and Prof. Andreas Geiger. Prior to that, I received my master's degree in robotics from ETH Zurich, and my bachelor's degree in electrical engineering from University of Stuttgart and Wuhan University.

Email  /  GitHub  /  Google Scholar  /  LinkedIn  /  Twitter

Publications

AG3D: Learning to Generate 3D Avatars from 2D Image Collections

Fast-SNARF: A Fast Deformer for Articulated Neural Fields

Vid2Avatar: 3D Avatar Reconstruction from Videos in the Wild via Self-supervised Scene Decomposition

InstantAvatar: Learning Avatars from Monocular Video in 60 Seconds

PINA: Learning a Personalized Implicit Neural Avatar from a Single RGB-D Video Sequence

gDNA: Towards Generative Detailed Neural Avatars

IM Avatar: Implicit Morphable Head Avatars from Videos

Human Performance Capture from Monocular Video in the Wild

Render In-between: Motion Guided Video Synthesis for Action Interpolation

SNARF: Differentiable Forward Skinning for Animating Non-Rigid Neural Implicit Shapes

Shape-aware Multi-Person Pose Estimation from Multi-view Images

Category Level Object Pose Estimation via Neural Analysis-by-Synthesis

Human Body Model Fitting by Learned Gradient Descent

Monocular Neural Image Based Rendering with Continuous View Control

Unpaired Pose Guided Human Image Generation

StreetMap - Mapping and Localization on Ground Planes using a Downward Facing Camera

Acknowledgement: Design and source code from Jon Barron's website and Leonid Keselman's Jekyll fork