Cloudminds Robotics Co., Ltd.

Xiao-Qing William Huang

ZhanChao Zhang

Bin Wang

Guanghua Yang

Shikui Ma

A robot control method includes: the first control instruction of the robot is obtained; the behavior blueprint corresponding to the first control instruction is determined, the behavior blueprint including the plurality of blueprint nodes connected according to the target sequence; and the preset functions corresponding to the blueprint nodes are executed according to the target sequence. The preset functions corresponding to the blueprint nodes may include the atomic API ability, the meta skill and the task skill. The atomic API ability is used for characterizing a specific function executed by the robot through a component, the meta skill is used for characterizing a function of an execution unit of the robot, the task skill is used for characterizing a function of user-oriented execution of a target task by the robot, the robot includes one or a plurality of execution units, and each execution unit includes one or more components.

Robot control method, storage medium and electronic device

An action learning method, including: acquiring human body moving image data; determining three-dimensional human body pose action data corresponding to the human body moving image data; matching the three-dimensional human body pose action data with atomic actions in a robot atomic action library to determine robot action sequence data corresponding to the human body moving image data; performing action continuity stitching on all robot sub-actions in the robot action sequence data sequentially; determining a continuous action learned by a robot from the robot action sequence data subjected to the action continuity stitching.

Growing community of inventors

Shanghai, China

Xiao-Qing William Huang