Information processing device, information processing method, and program

An information processing method implemented in an information processing device, the information processing method comprising performing:

• generating an environment map including type and layout information about an object in a real world;
• classifying an action object movable in the real world into a first action object representing a learning-processing target or a second action object representing a non-learning-processing target;
• generating first action analysis data of the first action object and second action analysis data of the second action object;
• generating correspondence data representing correspondence among the environment map, first time-series data of the first action analysis data, and second time-series data of the second action analysis data;
• generating an action model by performing a machine learning process using the correspondence data, wherein the action model is capable of outputting, using the environment map as an input value, a result of action estimation as an output value;
• and
• generating action control data for an agent, using the action model, wherein the agent includes at least one of a virtual character or a physical robot.

A non-transitory computer-readable storage medium having embodied thereon a program, which when executed by an information processing device of a computer causes the computer to execute a method, the method comprising:

• generating an environment map including type and layout information about an object in a real world;
• classifying an action object movable in the real world into a first action object representing a learning-processing target or a second action object representing a non-learning-processing target;
• generating first action analysis data of the first action object and second action analysis data of the second action object;
• generating correspondence data representing correspondence among the environment map, first time-series data of the first action analysis data, and second time-series data of the second action analysis data;
• generating an action model by performing a machine learning process using the correspondence data, wherein the action model is capable of outputting, using the environment map as an input value, a result of action estimation as an output value;
• and
• generating action control data for an agent, using the action model, wherein the agent includes at least one of a virtual character or a physical robot.