Penglei Sun

Fudan University, Hong Kong University of Science and Technology

Papers

Total Citations

H-Index

About

Penglei Sun is an emerging researcher working at the intersection of embodied AI, robotic manipulation, and multimodal knowledge representation. His work addresses a fundamental challenge in modern robotics: enabling agents to understand their environments intelligently and interact with the physical world in nuanced, human-aligned ways. Sun's most influential contribution, "Scene-Driven Multimodal Knowledge Graph Construction for Embodied AI" (2024, 38 citations), demonstrates his commitment to grounding AI systems in structured scene understanding — a critical capability for real-world robotic deployment. This work reflects his broader interest in building robots that can reason about their surroundings rather than simply react to them. A recurring theme in Sun's research is fine-grained robotic grasping. His work on 6-DoF part affordance grounding moves beyond object-level grasping toward more human-like, part-aware manipulation, a direction that has attracted growing attention across successive publications. His earlier work on human-in-the-loop grasping using BERT-based scene representations further highlights his dedication to natural language interfaces that make robots more accessible and responsive to human guidance. Collectively accumulating over 50 citations within a compact publication window, Sun's research trajectory positions him as a promising contributor to the next generation of intelligent, language-aware robotic systems.