Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – Investigators discuss new findings in androids. According to news originating from Hangzhou, People’s Republic of Chin a, by NewsRx correspondents, research stated, “Humanoid grasping is a critical a bility for anthropomorphic hand, and plays a significant role in the development of humanoid robots.” Our news reporters obtained a quote from the research from Zhejiang University: “In this article, we present a deep learning-based control framework for humanoi d grasping, incorporating the dynamic contact process among the anthropomorphic hand, the object, and the environment. This method efficiently eliminates the co nstraints imposed by inaccessible grasping points on both the contact surface of the object and the table surface. To mimic human-like grasping movements, an un deractuated anthropomorphic hand is utilized, which is designed based on human h and data. The utilization of hand gestures, rather than controlling each motor s eparately, has significantly decreased the control dimensionality. Additionally, a deep learning framework is used to select gestures and grasp actions. Our met hodology, proven both in simulation and on real robot, exceeds the performance o f static analysis-based methods, as measured by the standard grasp metric Q1.”
NETL
NSTL