Di (Kilian) Chang

常迪

CS Ph.D. @ University of Southern California

Student Researcher @ TikTok - Seed Vision Research

Department of Computer Science
Viterbi School of Engineering
University of Southern California
Email1: dichang at usc dot edu (primary)
Email2: dchang at ict dot usc dot edu

    [News] [Education] [Publications] [Experience]


About Me

                               

I am a first-year second-year third-year Ph.D. student in the Department of Computer Science at University of Southern California (USC) with Prof. Mohammad Soleymani of IHP-Lab. I'm also a student researcher at TikTok-ByteDance Seed Vision Research, working with Dr. Yichun Shi and Dr. Peng Wang. During my PhD at USC, I'm fortunate to have collaborated with Prof. Gordon Wetzstein from Stanford University, Prof. Narendra Ahuja from University of Illinois Urbana-Champaign, and their students. I was a Research Scientist Intern at TikTok-ByteDance Intelligent Creation Lab, working with Dr. Hongyi Xu and Dr. Linjie Luo during 2024 summer, and with Dr. Yichun Shi and Dr. Xiao Yang during 2023 summer.

Before that, I spent four wonderful years during my undergraduate at Dalian University of Technology and Technical University of Munich, studying Informatics and Information Engineering. I'm honored to be advised by Prof. Matthias Niessner and Dr. Aljaž Božič during my praktikum at Visual Computing Group. I spent a wonderful summer in 2022 with Dr. Tong Zhang and Prof. Sabine Süsstrunk at EPFL CS and in 2021 with Prof. Dan Xu at HKUST CSE.

I'm interested in Generative Vision Models, 3D Reconstruction and Motion Synthesis. My current and past research focuses include:

Research opportunities:
For USC undergraduate students: We have openings for interns through CURVE program. Usually we recruit students (first-time researchers and continuing researchers) during Fall and Spring semesters, please apply early.

For USC master students: Please directly email Prof. Mohammad Soleymani for such inquiries. It would be better if you've attended CSCI 535 and obtained a good grade.

For all other students outside USC: we usually don't have openings during Fall and Spring semesters, but Prof. Mohammad Soleymani sometimes recruits good candidates (Undergrad/Ms/visiting PhD) during the Summer. Email him for further details.

I am happy to collaborate and/or answer questions about my research and my previous study at TUM and current PhD program at USC. If you are interested in research collaboration or have any inquiries about my experience, please send me an email.

News


Talks

EPFL School of Computer and Communication Sciences - 09/26/2024

Preprints

            
DiTaiListener: Controllable High Fidelity Listener Video Generation with Diffusion

Masksim Siniukov*, Di Chang*, Minh Tran, Hongkun Gong, Ashutosh Chaubey, Mohammad Soleymani
(*equal contribution)
Arxiv Preprint

[paper]

X-Dancer: Expressive Music to Human Dance Video Generation

Zeyuan Chen, Hongyi Xu, Guoxian Song, You Xie, Chenxu Zhang, Xin Chen, Chao Wang, Di Chang, Linjie Luo
Arxiv Preprint

[paper]

MagicPose4D: Crafting Articulated Models with Appearance and Motion Control

Hao Zhang*, Di Chang*, Fang Li, Mohammad Soleymani, Narendra Ahuja
(*equal contribution   equal advise)
Arxiv Preprint

[project page] [paper] [code] GitHub stars

Learning-based Multi-View Stereo: A Survey

Fangjihua Wang*, Qingtian Zhu*, Di Chang*, Quankai Gao, Junlin Han, Tong Zhang, Richard Hartley, Marc Pollefeys
(*equal contribution   project lead)
Arxiv Preprint

[paper]


Selected Publications [Google Scholar]

                  
X-Dyna: Expressive Dynamic Human Image Animation

Di Chang, Hongyi Xu*, You Xie*, Yipeng Gao*, Zhengfei Kuang*, Shengqu Cai*, Chenxu Zhang*, Guoxian Song, Chao Wang, Yichun Shi, Zeyuan Chen, Shijie Zhou, Linjie Luo, Gordon Wetzstein, Mohammad Soleymani,
(*equal contribution as second authors)
IEEE/CVF Computer Vision and Pattern Recognition Conference (CVPR), 2025.

[project page] [paper] [code] GitHub stars

DIM: Dyadic Interaction Modeling for Social Behavior Generation

Minh Tran*, Di Chang*, Maksim Siniukov, Mohammad Soleymani
(*equal contribution)
European Conference on Computer Vision (ECCV), 2024.

[project page] [paper] [code] GitHub stars

MagicPose: Realistic Human Poses and Facial Expressions Retargeting with Identity-aware Diffusion.

Di Chang, Yichun Shi, Quankai Gao, Jessica Fu, Hongyi Xu, Guoxian Song, Qing Yan, Yizhe Zhu, Xiao Yang, Mohammad Soleymani
International Conference on Machine Learning (ICML), 2024.

[project page] [paper] [code] [video] GitHub stars

DiffPortrait3D: Controllable Diffusion for Zero-Shot Portrait View Synthesis

Yuming Gu, Xie You, Hongyi Xu, Guoxian Song, Yichun Shi, Di Chang, Jing Yang, Linjie Luo
IEEE/ CVF International Conference on Computer Vision and Pattern Recognition (CVPR), 2024. (Highlight)

[project page] [paper] [code] [video] GitHub stars

LibreFace: An Open-Source Toolkit for Deep Facial Expression Analysis

Di Chang, Yufeng Yin, Zongjian Li, Minh Tran, and Mohammad Soleymani
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2024. (Application Track)

[project page] [paper] [code] [video]

RC-MVSNet: Unsupervised Multi-View Stereo with Neural Rendering

Di Chang, Aljaž Božic, Tong Zhang, Qingsong Yan, Yingcong Chen, Sabine Süsstrunk and Matthias Nießner
European Conference on Computer Vision (ECCV), 2022.

[project page] [paper] [code] [video] GitHub stars


Education

University of Southern California, Los Angeles, USA
(Aug. 2022 - Present)

  • Doctor of Philosophy in Computer Science
  • Major Orientation: Deep Learning for 3D Vision and Affective Computing

Technical University of Munich, Munich, Germany (Aug. 2021 - Jul. 2022)

  • Bachelor of Science in Informatics
  • Major Orientation: Deep Learning for 3D Perception, 3D Scanning and Spatial Learning
  • Overall GPA: 1.2/1.0
  • Bachelor thesis: "Supervised and Unsupervised Multi View Stereo for Depth Inference"

Dalian University of Technology, Dalian, China (Sep. 2018 - Jul. 2021)

  • Bachelor of Engineering in Eletronic Information Engineering
  • Major Orientation: Object Detection and Tracking
  • Overall GPA: 91.5/100 | 3.93/4.0

Intern & Work Experience

TikTok, Los Angeles, CA, USA (Dec. 2024 - Current)

  • Position: Student Researcher at Seed Vision Research
  • Supervisor: Dr. Yichun Shi , and Dr. Peng Wang
  • Project: Foundation Models for Text-guided Image Editing and Stylization

TikTok, San Jose, CA, USA (May. 2024 - Nov. 2024)

TikTok, San Jose, CA, USA (May. 2023 - Nov. 2023)

EPFL, Lausanne, Switzerland (Jul. 2022 - Oct. 2022)

  • Position: Summer@EPFL (Top 2% from the candidates in the world) hired by IVRL
  • Supervisor: Dr. Tong Zhang and Prof. Sabine Süsstrunk
  • Project: Video Synthesis with Diffusion Models

TUM, Munich, Bayern, Germany (Aug. 2021 - Mar. 2022)

HKUST, Hong Kong, China (Mar. 2021 - Sep. 2021)

Students


Academic Service

Reviewer of the following conferences:

ECCV 2022, 2024
NeurIPS 2022, 2024
CVPR 2024, 2025
ICLR 2025
ICML 2025
ICCV 2025

Teaching


Miscellaneous

Camaro SS

This page has been visited for HTML Hit Countertimes

Last update: Mar. 10, 2025