Biography

I am currently a 2nd-Year Ph.D. student in Computer Science, working in Perceiving Systems, Max Planck Institute for Intelligent Systems, with Michael J. Black.

I was briefly enrolled (2019-2020) as a Ph.D. student in Vision and Graphics Lab at University of Southern California before I decided to relocate at Tuebingen, advised by Hao Li (Now the CEO and Co-founder of Pinscreen). I got M.Sc. degree at Shanghai Jiao Tong University in 2019, working in Machine Vision and Intelligence Group, advised by Cewu Lu, and got B.Eng. degree in Digital Media Technology at Shandong University in 2016, worked closely with Lu Wang.

I also spent wonderful time at USC Institute for Creative Technology with Jun Xing (Now the lead researcher of miHoYo), Kwai Y-tech Graphics AI Teamwith Chongyang Ma, LightChaser Animation Studio with Xinxin Zhang (Now the co-founder and CTO of Zenus), and is working as R&D intern at Ubisoft La Forge.

I am a Virtual Avatar Researcher, playing with pixels and voxels. My research interest mainly lies in the intersection of computer vision and computer graphics, especially in 3D Human Digitalization (i.e., Reconstruction, Modeling, Animation) and 2D Human Pose Estimation and Tracking.

News

[2022/04/20] Invited to talk about ICON at Talking Papers Podcast. Great chat with Yizhak Ben-Shabat!
[2022/04/07] 走进马克思普朗克智能系统研究所与苏黎世联邦理工AIT团队 视频(中文) [上, 下].
[2022/03/02] ICON got accepted by CVPR 2022 → Reader Discord 
[2022/02/01] I am invited to give Tech Talks by USC-ICT, and Meta Reality Labs, Pittsburgh (PDF, 21MB).
[2021/12/17] New work "ICON: Implicit Clothed humans Obtained from Normals", please check its page.
---- show more ----

Publications

ICON: Implicit Clothed humans Obtained from Normals 

Yuliang Xiu, Jinlong Yang, Dimitrios Tzionas, Michael J. Black
Conference on Computer Vision and Pattern Recognition 2022 (CVPR 2022)

Intro  Home  [arXiv, Reader] Slides  Video [En, 中]  Zhihu  Bibtex  

                
              		
            


            

              
                
              
              

                Monocular Real-Time Volumetric Performance Capture 


                Yuliang
                  Xiu*, Ruilong
                  Li*, Shunsuke
                  Saito, Zeng Huang,
                Kyle Olszewski, Hao Li

                (*equal contribution)
                

                European Conference on Computer Vision (ECCV 2020)
                


                Intro 
                Home 
                [arXiv , Reader]

                Video 
                 Zhihu 
                Bibtex
                   
                  
                  

                  
                		
              


              

                
                  
                
                
                  Volumetric Human Teleportation 


                  Ruilong Li,
                  Kyle Olszewski,
                  Yuliang
                    Xiu, Shunsuke
                    Saito, Zeng Huang,
                  Hao Li

                  

                  SIGGRAPH Real-Time Live
                      2020 (Best
                      in Show Award)
                  


                  Intro 
                  Home 
                  [Paper ,
                  Reader]
                  Video 
                   Media (
                    1,
                    2,
                    3,
                    4) 
                  Bibtex
                     
                    
                    
                    
                  		
                


                

                  
                    
                  
                  

                    Pose Flow: Efficient Online Pose Tracking 


                    Yuliang Xiu,
                    Jiefeng Li, Haoyu Wang,
                    Yinghong
                    Fang, Cewu Lu

                    

                    British Machine Vision Conference (BMVC
                        2018)
                    


                    Intro 
                    [arXiv ,
                    Reader]
                     Zhihu 
                    Bibtex
                       
                      
                      
                      
                    		
                  

                  

                    
                      
                    
                    

                      PoseHD: Boosting Human Detectors using Human Pose
                        Information


                      Zhijian Liu, Bowen
                        Pan,
                      Yuliang Xiu, Cewu
                        Lu
                      

                      AAAI Conference on Artificial Intelligence (AAAI 2018)
                      


                      Intro 
                      Paper 

                      Bibtex
                         
                        
                        
                        
                      		
                    

                  


                  

                  
Talks

                    
                    走进马克思普朗克智能系统研究所与苏黎世联邦理工AIT团队 (Part A (2h), Part B (2h))


                   
                      ICON: 提高三维数字人重建的姿势水平 (Keynote (726MB),
                        Video (60
                        min))


                      Towards Large-scale Avatar Creation From In-the-wild
                        Pixels (PDF,
                      21MB)
                      
                      Pose trajectory extraction and novel-view synthesis
                        from visual content (Link)
                      
                    

                    
Service

                          Journal: T-PAMI / SIGGRAPH 2022

                          Conference: CVPR 2022 / ECCV 2022

                    
Blogs

                        [翻译]当我们谈科学研究的创新性时，我们在谈些什么 Source 
                        译文 


                    
                    
                    

                      

                        
                        
                      


                      This page is last updated at