About Me
I am currently building a team focused on interactive video and world models. If you share an interest in this direction, please feel free to reach out directly.
Before that, I led the content generation team for the Qwen App at Alibaba Group, building systems that connect AI agents, image generation, and interactive video generation into a single intelligent creation experience.
My work sits at the intersection of generative models and real-world products. Over the past years I have focused on several closely related topics: controllable diffusion models, subject-driven personalization, human-centric video and talking-avatar generation, visual editing (makeup / hair / garment / font / face), efficient inference, and agentic creation workflows.
Earlier, I was Head of AI at Tiamat AI, where I built the full-stack generative pipeline for a virtual-companion app; led a AIGC group at Xiaohongshu to power content creation; and led the GAN team at Baidu on intelligent font production and virtual-human technologies after starting in Baidu IDL's OCR team. Several works have been widely adopted by the open-source community, including LiveAvatar (Hugging Face #1 Paper of the Day, 2k+ GitHub stars) and EasyControl (Hugging Face #1 trending Space).
News
- Jun 2026 LiveAvatar and CollectionLoRA accepted to ECCV 2026.
- 2026 RationalRewards released — Hugging Face #2 Paper of the Day and accepted to COLM 2026 . EasyText published at AAAI 2026; Stable-Hair V2 published in IEEE TVCG.
- Dec 2025 LiveAvatar released — a 14B-parameter real-time streaming avatar model running at 45 FPS on 5×H800. Hugging Face #1 Paper of the Day, 2k+ GitHub stars.
- Oct 2025 Three papers presented at ICCV 2025: EasyControl, Any2AnyTryOn, and Fonts.
- Aug 2025 StableMakeup presented at SIGGRAPH 2025.
- Apr 2025 EasyControl Ghibli became the #1 trending Space on Hugging Face.
- Feb 2025 Stable-Hair published at AAAI 2025.
- Feb 2023 Win the champion at Xiaohongshu's 3rd Hackathon
- Jul 2021 Win the champion at Baidu's 25th Hackathon
Selected Publications
European Conference on Computer Vision (ECCV)
2025 IEEE/CVF International Conference on Computer Vision (ICCV)
2025 IEEE/CVF International Conference on Computer Vision (ICCV)
Proceedings of the AAAI Conference on Artificial Intelligence
Proceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers
2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Experience
- 2025 – Present
- Tech Lead, Qwen App, Alibaba Group
- 2024 – 2025
- Head of AI, Tiamat AI
- 2022 – 2024
- AIGC Lead, Xiaohongshu
- 2019 – 2022
- Tech Lead, GAN Team, Baidu, Vision Technology Department
- 2017 – 2019
- Research Engineer, OCR Team, Baidu IDL
Honors & Awards
- Silver Medal, China National Patent Award
- Champion, Baidu 25th Hackathon (face-swap virtual idol group)
- Champion, Xiaohongshu 3rd Hackathon
- Hugging Face #1 Paper of the Day (LiveAvatar), #2 Paper of the Day (RationalRewards)
- Hugging Face Trending Spaces (EasyControl-ghibli)