Perceiving, Reasoning, Adapting: A Dual-Layer Framework for VLM-Guided Precision Robotic Manipulation

Qingxuan Jia, Guoqin Tang, Zeyuan Huang +5

Vision-Language Models (VLMs) demonstrate remarkable potential in robotic manipulation, yet challenges persist in executing complex fine manipulation tasks with high speed and prec…