Xiao Li (李潇)

Xiao Li (李潇)

Researcher in CG & CV

Microsoft Research Asia

Bio

I am currently a senior researcher in Media Computing group at Microsoft Research Asia (MSRA). Prior to joining MSRA, I was a researcher / alogrithm engineer in Graphics AI team at Kuaishou Technology (Kwai Inc.).

I got my Ph.D. degree from the joint Ph.D. program between Unitersity of Science and Technology of China (USTC) and MSRA in 2019, under the supervision of Dr. Baining Guo and Dr. Yong Wang. I obtained my B.Eng. degree in Eletronic Engineering also from USTC in 2013.

From 2014 to 2019, I was a long-term research intern in Internet Graphics group at MSRA, where I conducted my Ph.D. research with thesis topic on appearance and geometry modeling with deep learning approach.

(Here is an article about my wonderful internship experience at MSRA.)

Currently my research interests fall in computer graphics and computer vision. Specifically, I am interested in 3D synthesis and generation, neural rendering, and multi-modality.

Internship chances (Updated on 03/08/2023): Our group is looking for research interns on computer vision, speech and video codec. Please drop me an email if you are interested in. For gap-year students, please check the MSRA StarBridge program if you want to apply.

Interests

  • Appearance modeling
  • 3D reconstruction and generation
  • Differentible / Neural Rendering

Education

  • Ph.D. in Pattern Recognition & Intelligent Systems, 2019

    University of Science and Technology of China

  • B.Eng. in Electronic Engineering, 2013

    University of Science and Technology of China

Publications

(*: Interns at MSRA)

Unsupervised Temporal Correspondence Learning for Unified Video Object Removal

IEEE Transactions on Image Processing

Efficient View Synthesis with Neural Radiance Distribution Field

ICCV 2023

R^2-VOS: Robust Referring Video Object Segmentation via Relational Cycle Consistency

ICCV 2023

High-Fidelity and Freely Controllable Talking Head Video Generation

CVPR 2023

SCoDA: Domain Adaptive Shape Completion for Real Scans

CVPR 2023

Structural Multiplane Image: Bridging Neural View Synthesis and 3D Reconstruction

CVPR 2023

Two-shot Video Object Segmentation

CVPR 2023

Neural Capture of Animatable 3D Human from Monocular Video

ECCV 2022

Video Instance Segmentation by Instance Flow Assembly

IEEE Transactions on Multimedia

Distribution Aligned Multimodal and Multi-Domain Image Stylization

ACM Transactions on Multimedia Computing, Communications, and Applications

Deep Inverse Rendering for High-resolution SVBRDF Estimation from an Arbitrary Number of Images

SIGGRAPH 2019

Capturing Piecewise SVBRDFs with Content Aware Lighting

CGI 2019 (Short paper)

Mimicking the In-Camera Color Pipeline for Camera-Aware Object Compositing

arxiv Preprint

Single Photograph Surface Appearance Modeling with Self-augmented CNNs and Inexact Supervision

Pacific Graphics 2018

Hair Interpolation for Portrait Morphing

Pacific Graphics 2013