Biography
I am a Research Scientist/Engineer II at Adobe Research. My work spans physically based simulation, geometry processing, 3D data generation, generative AI, and embodied AI. I am interested in building scalable, physically grounded systems for creating, understanding, and interacting with 3D worlds, with applications in content creation, spatial intelligence, video/world models, robotics, and agentic creative workflows.
At Adobe, I have worked on physics-based modeling and real-time animation systems, including physics-driven animation technologies used by millions of users in Adobe Express. More recently, my work has focused on scalable 3D synthetic data generation, dynamic scene construction, and model-facing data pipelines for generative and spatial AI systems.
Prior to joining Adobe, I received my Ph.D. in Computer Science from the University of Texas at Austin, supervised by Prof. Etienne Vouga. Before that, I received my B.S. in Information and Computing Science from the University of Science and Technology of China in 2018, where I worked with Prof. Ligang Liu.
Ph.D. Dissertation: Complex Wrinkle Simulation and Robust Surface Remeshing | UT Austin Library
Download my resumé
- Physical Simulation
- Geometry Processing
- Generative AI
- Embodied AI
Ph.D in Computer Science, 2018 - 2024
The University of Texas at Austin
B.S. in Information and Computing Science, 2014 - 2018
University of Science and Technology of China
Research
Experience
- 3D Synthetic Data & Generative AI: Developing scalable 3D data generation workflows for synthetic scenes, dynamic content, camera/view control, and model-facing data pipelines for generative and spatial AI systems.
- Animation Systems (Adobe Express): Led development and launch of real-time physics-driven animation systems, enabling scalable and interactive motion effects used by millions of users.
- Simulation Systems: Designed and deployed high-performance simulation systems for deformable objects, with a focus on robustness, numerical stability, and scalability in production environments.
- AI-Driven Simulation: Built simulation backends integrated with learning-based models to generate consistent geometry and motion for large-scale content creation workflows.
- Generative Pipeline: Developed pipelines combining simulation and rendering for scalable 3D content generation and asset creation.
- Developed robust low-poly mesh generation techniques for general 3D models, handling challenging cases such as small gaps, holes, and non-manifold geometry.
- Built reliable geometry processing pipelines for real-world assets, emphasizing robustness, efficiency, and scalability.
- Developed stable time integration schemes for physically-based simulation under challenging contact scenarios.
Project Demos
Adobe Express — Dynamic Animation
Flying Rabbit
Bounce Cat
Bounce
Clouds