About Me

Hi, there! 👋 I’m glad you found my website.

I’m currently…

🎓 Pursuing a degree in Computer Science at ShanghaiTech University.
🌱 Engaged in exploring the fields of Machine Learning and Computer Graphics.
🌐 Expanding my knowledge in Web Development, with hands-on experiences in React and Vue.
📷 A photography and videography enthusiast.
🎈 An avid participant in programming contests, with a background in OI and ACM contests.
💼 Interning at Deemos, contributing to advancements in DigitalHuman and 3D face reconstruction.
🌉 Participated in the exchange program at the University of California Berkeley, College of Engineering
💡 Learn more about my work and research on my website.

skillicons

Built a Qt application using ADB to collect, analyze, and visualize data from drones, enhancing data insights and decision-making for stakeholders.
Developed a real-time visualization analyzer for collecting and analyzing frequency scan data (2.4Ghz, 5.2Ghz, and 5.8Ghz) from drones, leading to more accurate data interpretation.
Employed parallel computing techniques to accelerate data processing and visualization, achieving more efficient data analysis and a faster feedback loop.
Profiled and optimized the existing system, resulting in enhanced application performance and reduced processing time for data analysis.

Created a real-time 3D interactive avatar system demo utilizing audio-driven facial expression technologies, showcased at GAIDC (Global AI Developer Conference 2023) to thousands of attendees.
Led the development of the Hyperhuman frontend, a web-based application on the HYPERHUMAN platform that enables users to create, edit, animate, and stylize 3D digital humans. Features include:
- Generation and editing of 3D digital avatar models with high-fidelity PBR textures using a prompt.
- Creation and modification of 3D digital avatar models from a single photo, featuring high-fidelity PBR textures.
- Automatic binding and RGB-video-based driving using extensive 4D and expression data. (Reference: Zhang, Longwen, et al. “Video-driven Neural Physically-based Facial Asset for Production.” SIGGRAPH Asia 2022)
- Support for multilingual audio-driven 3D digital avatars.
- User-friendly stylization tool for 3D digital humans.
- Automatic topology conversion.
View the demo at Hyperhuman Demo on Twitter.
Or view our current version of Hyperhuman

Led the creation of the Freshmen Handbook for our university.
The driving force behind CourseBench, a platform for students to discuss and review their courses.
Organizer of numerous workshops and lectures, including “DJI Workshop at ShanghaiTech” and “SI100B Beginner’s Guide.”
Developed a variety of tools and games: G22G Online Game, CoursePrettier, and GPT Refinement System.