Xihe Qiu

Shanghai University of Engineering Science

Papers

Total Citations

H-Index

About

Xihe Qiu is a researcher at the forefront of embodied AI and robotic perception, with key contributions spanning 6D pose estimation, multi-agent reinforcement learning, and human-robot collaboration. Their work on EFN6D (2022, 9 citations) introduced an efficient RGB-D fusion network that significantly advanced object pose estimation for robotic manipulation, achieving robust performance under challenging lighting and occlusion conditions. More recently, Qiu has pioneered the integration of large language models (LLMs) with reinforcement learning, as demonstrated in their LMGT framework (2025, 5 citations), which uses LLM-generated reward signals to guide agents in complex tasks—a breakthrough for sample efficiency and task transfer. Their 2025 work on multiagent fuzzy reinforcement learning for endovascular robotics (8 citations) addresses the critical challenge of cooperative navigation in surgical settings, enabling autonomous coordination of guidewires and catheters through tortuous vascular paths. This research has direct implications for minimally invasive surgery, reducing operator cognitive load while improving precision. With a growing citation footprint and a focus on bridging perception, language, and control, Qiu is shaping the next generation of intelligent robotic systems for both industrial and medical applications.