Shen Sang (桑燊)

I am a research engineer, with a primary focus on virtual human, at the Intelligent Creation Lab, TikTok.

I received my master's degree from UC San Diego. I was fortunate to work with Prof. Manmohan Chandraker at the Visual Computing Center.

Github  /  Google Scholar  /  LinkedIn

Please feel free to contact me by email: shen.sang[at]bytedance[dot]com

profile photo

[2024.02] TikTok AI Self is launched and will be available globally soon. stay tuned!

[2023.07] ActorsNeRF is accepted to ICCV 2023.

[2023.05] TikTok AI Avatars is launched globally.

[2022.08] AgileAvatar is accepted to SIGGRAPH Asia 2022.

[2022.06] TikTok Avatars is launched globally.


I work on avatar creation, focusing on many aspects of diffusion models now. My current interests include diffusion model personalization (see TikTok AI Moji, TikTok AI Self and TikTok AI Avatars) and training and inference efficiency of diffusion models. Previously, I also worked on 3D vision and graphics, particularly in 3D reconstruction and neural relighting.

ActorsNeRF: Animatable Few-shot Human Rendering with Generalizable NeRFs
Jiteng Mu, Shen Sang, Nuno Vasconcelos, Xiaolong Wang

ICCV, 2023
AgileAvatar: Stylized 3D Avatar Creation via Cascaded Domain Bridging
Shen Sang, Tiancheng Zhi, Guoxian Song, Minghao Liu, Chunpong Lai, Jing Liu, Xiang Wen, James Davis, Linjie Luo

SIGGRAPH Asia, 2022
arXiv / paper / bibtex
OpenRooms: An Open Framework for Photorealistic Indoor Scene Datasets
Zhengqin Li, Ting-Wei Yu, Shen Sang, Sarah Wang, Meng Song, Yuhan Liu, Yu-Ying Yeh, Rui Zhu, Nitesh Gundavarapu, Jia Shi, Sai Bi, Zexiang Xu, Hong-Xing Yu, Kalyan Sunkavalli, Miloš Hašan, Ravi Ramamoorthi, Manmohan Chandraker

CVPR, 2021   (Oral Presentation)
Single-Shot Neural Relighting and SVBRDF Estimation
Shen Sang, Manmohan Chandraker

ECCV, 2020
paper / bibtex
TikTok AI-Moji, 2024
TikTok AI Self, 2024
TikTok AI Avatars, 2023
TikTok Avatars, 2022

Paper review for: CVPR 2024, AAAI 2024, SIGGRAPH Asia 2023, ICCV 2023, CVPR 2023, ICLR 2023, AAAI 2023, TMM, C&G

Credits to this template.