Max Robotics — Global Service Robot Directory + US Certification

About

Qiwei Meng is a rising researcher at the forefront of integrating large language models (LLMs) with robotic perception and manipulation. Their work primarily focuses on long-horizon task planning, 6D object pose estimation, and human-robot interaction. Meng’s most impactful contribution is the FLTRNN framework (2024, 26 citations), which addresses the critical challenge of faithful long-horizon planning by enhancing LLM-based planners with recurrent neural networks to maintain execution correctness over extended tasks. In object perception, Meng developed KGNet (2023, 15 citations), a knowledge-guided network that enables category-level 6D pose and size estimation without requiring exact CAD models—a key enabler for robotic grasping in unstructured environments. Additional contributions include multi-scale graph convolutional networks for 3D human pose estimation (MSMB-GCN) and geometric primitive deformation for unseen object pose estimation (GPD). Meng’s work on decision-making in robotic grasping with LLMs and referring expression comprehension in semi-structured human-robot interaction further demonstrates their commitment to bridging language understanding and physical robot control. With a rapidly growing citation record and a clear trajectory toward more autonomous, language-guided robotics, Qiwei Meng is establishing themselves as a notable voice in embodied AI and robot learning.