Relating to predicting the motion (s) of the object (s) in the robotic environment based on the image (s) capturing the object (s) and parameter (s) for future robot motion in the environment Methods and apparatus

摘要

Some embodiments of the present disclosure generally relate to a dip machine learning method and apparatus for predicting motion (s) (if any) to occur with respect to object (s) in an environment of a robot in response to a particular motion of the robot . Some implementations use a deep neural network model to predict at least one transformation (if any) of an image of the environment of the robot that occurs as a result of implementing at least some of the robot's specific movements in the environment / RTI The trained deep neural network model can predict a transformation based on an input that includes an image and a group of robot motion parameters that define a portion of a particular motion.