Hi, I am Hanwei Liu (刘汉伟, Liu Hanwei), a researcher at Huawei (华为) and a member of Huawei’s TopMind program. My current research focuses on World Models, Physical AI, and 3D/4D scene representation and rendering. I am particularly interested in building world models that bridge photorealistic rendering and physical dynamics, enabling realistic and physically plausible prediction, interaction, and simulation of the real world.
I am currently working on AI applications powered by Huawei MetaEngine, a cloud-native intelligent rendering engine, and MetaStudio, Huawei’s digital human platform. My work includes 3D/4D scene representation and rendering, digital human generation, generative world models, and physically grounded simulation for autonomous driving and embodied intelligence.
I received my Ph.D. degree in 2025 from the School of Electronics and Information Engineering, Tongji University, where I was supervised by Prof. Hui Xiao and Prof. Xuefeng Li. During my Ph.D., I worked as a research intern at NetEase Fuxi AI Lab(Norface: first unified top-1 across multiple FER/AU tasks, collaborated on DiffSFSR text-to-image generating 135 facial expressions), and Happy Elements(AI-assisted game content: 2D/3D lip-sync & facial retargeting, 3D motion & retargeting, 3D content generation), where I had the opportunity to collaborate with Yu Ding, former Chief AI Digital Human Expert at NetEase Fuxi.
My technical trajectory is: facial expression recognition and affective computing → digital human generation and animation → 3DGS/XDGS reconstruction and photorealistic rendering → 3D/4D World Models and Physical AI.
💬 liuhw01 AT gmail DOT com
🔥 News
- 2026:🎉🎉 I served as a reviewer for ECCV 2026.
- 2025:🎉🎉 I joined Huawei as a Researcher and was selected for Huawei’s TopMind program.
- 2025:🎉🎉 Our work on multi-modal expressive personality recognition was made public.
- 2025:🎉🎉 I served as a reviewer for CVPR 2025 and ACM MM 2025.
- 2024:🎉🎉 Our Virtual Human Group won all five championships in the 6th ABAW Competition at CVPR 2024. 🏆
- 2024:🎉🎉 Our work Norface was accepted by ECCV 2024.
- 2024:🎉🎉 Our work DiffSFSR was accepted by CVPR 2024.
- 2023:🎉🎉 Our work DUML was accepted by ACM MM 2023.
📝 Selected Publications [Full List]
* for the corresponding author.
- Happyface: High-Fidelity Audio-Driven Talking Face Generation.
- Under review
- Multi-modal expressive personality recognition in data non-ideal audiovisual based on multi-scale feature enhancement and modal augment.
- Weixuan Kong, Jinpeng Yu, Zijun Li, Hanwei Liu, Jiqing Qu, Hui Xiao, Xuefeng Li
- Under review
- Norface: Improving Facial Expression Analysis by Identity Normalization.
- Hanwei Liu, Rudong An, Zhimeng Zhang, Bowen Ma, Wei Zhang, Yan Song, Yujing Hu, Wei Chen, and Yu Ding
- The 18th European Conference on Computer Vision (ECCV 2024)
- Learning from more: Combating uncertainty cross-multidomain for facial expression recognition
- Hanwei Liu, Huiling Cai, Qingcheng Lin, Xuefeng Li, Hui Xiao
- Proceedings of the 31st ACM International Conference on Multimedia (ACMMM 2023)
- A review of intelligent music generation systems
- Lei Wang, Ziyi Zhao, Hanwei Liu, Junwei Pang, Yi Qin, Qidi Wu
- Neural Computing and Applications
- Adaptive multilayer perceptual attention network for facial expression recognition
- Hanwei Liu, Huiling Cai, Qingcheng Lin, Xuefeng Li, Hui Xiao
- IEEE Transactions on Circuits and Systems for Video Technology (IEEE TCSVT)
📖 Educations & Work Experience
- *2025 – *, Researcher, Huawei, Inc. (huawei.com).
- 2024 – 2025, Research Intern, Happyelements, Inc. (happyelements.com).
- 2023 – 2024, Research Intern, NetEase, Inc. (netease.com).
- 2020 – 2025, PhD student, Tongji University (tongji.edu.cn).
- 2017 – 2020, MS student, Nanjing University of Science and Technology (njust.edu.cn).
📫 Services
- Conference reviewer: ECCV 26, ACMMM 24&25, CVPR 25
- Journal Reviewer: IEEE Transactions on Image Processing, IEEE Transactions on Circuits and Systems for Video Technology, IEEE Transactions on Multimedia, IEEE Transactions on Fuzzy Systems, IEEE Transactions on Affective Computing, IEEE Transactions on Industrial Informatics, IEEE Signal Processing Letters, IEEE Access, IEEE/CAA Journal of Automatica Sinica