About
I’m a Ph.D student in Shanghai Jiao Tong University (SJTU), under the supervision of Prof. Cewu Lu. Prior to that, I received my Bachelor’s degree from SJTU in 2022.
My research focuses on how agents interact with objects. Prior to 2024, I concentrated on Human-Object Interaction (HOI) — detecting and recognizing interaction semantics between humans and objects, with representative works including PartMap, Symbol-LLM, and Pangea.
Since then, I have expanded my research toward transferring HOI knowledge to robotic systems, and GPS represents an initial outcome of this ongoing effort.
Experience
Research Intern — ByteDance
Seed Model Beijing, China
Embodied Multimodal Foundation Models
Selected Publications
GPS: Geometric Primary Structure for Articulated Parts Perception in Robot Manipulation
VR-based data collection system for articulated objects in real-world, which balances scalability and data quality.
From Isolated Islands to Pangea: Unifying Semantic Space for Human Action Understanding
A unified framework that bridges isolated action datasets by learning shared semantic representations across different human activity domains.
Symbol-LLM: Leverage Language Models for Symbolic System in Visual Human Activity Reasoning
Leverage Large Language Models and neuro-symbolic computation to generate human part primitives, thus improving activity reasoning.
Mining Cross-Person Cues for Body-Part Interactiveness Learning in HOI Detection
Human-Object Interaction learning from a global, scene-level perspective.
HAKE: A Knowledge Engine Foundation for Human Activity Understanding
Program interpretable logic rules to infer human activity semantics.
Transferable Interactiveness Knowledge for Human-Object Interaction Detection
Introduce transferable interactiveness knowledge to enhance human-object interaction detection.
Selected Awards
- Huatai Securities Technology Scholarship (2025)
- Shanghai Outstanding Graduates (2022)
- National Scholarship (2019)