Abstract: A learning method explores, in a block space, a global path from a sub initial point to a sub goal candidate region for movement of an agent, and limits, based on the global path, an exploring space to thereby determine a limited space in the exploring space. The method arranges a sub goal in the limited space in accordance with a position of a goal point, and transforms absolute coordinates of each of at least one obstacle and a sub goal in the limited space into corresponding relative coordinates relative to a position of an agent located in the limited space. Then, the method explores, in the limited space, a target path from the initial point to the sub goal.

Abstract: In a trajectory generation apparatus, position coordinates of an obstacle existing in a motion space of a robot arm is acquired. A hand position at a second time, which is a time next to a first time, is estimated by using a learning result of machine learning, based on the position coordinates of the obstacle, a subject joint state of the robot arm at the first time, and a target joint state of the robot arm. A non-interfering joint state of the robot arm at which the obstacle does not interfere with the robot arm at the second time is searched for by using the hand position as a restriction.

Abstract: In a trajectory generation apparatus, position coordinates of an obstacle existing in a motion space of a robot arm is acquired. A hand position at a second time, which is a time next to a first time, is estimated by using a learning result of machine learning, based on the position coordinates of the obstacle, a subject joint state of the robot arm at the first time, and a target joint state of the robot arm. A non-interfering joint state of the robot arm at which the obstacle does not interfere with the robot arm at the second time is searched for by using the hand position as a restriction.

Abstract: A learning method explores, in a block space, a global path from a sub initial point to a sub goal candidate region for movement of an agent, and limits, based on the global path, an exploring space to thereby determine a limited space in the exploring space. The method arranges a sub goal in the limited space in accordance with a position of a goal point, and transforms absolute coordinates of each of at least one obstacle and a sub goal in the limited space into corresponding relative coordinates relative to a position of an agent located in the limited space. Then, the method explores, in the limited space, a target path from the initial point to the sub goal.

Abstract: An arithmetic processing apparatus performs arithmetic by a neural network in which multiple processing layers are hierarchically connected. The arithmetic processing apparatus corresponding to one of the multiple processing layers includes a convolution arithmetic portion and a pooling processing portion. The convolution arithmetic portion receives an input data from another of the plurality of processing layers, performs convolution arithmetic to the input data, and in each arithmetic cycle, outputs a part of all convolution arithmetic result data required for single pooling processing. The pooling processing portion performs the single pooling processing to the all convolution arithmetic result data before executing activation processing.

Abstract: An arithmetic processing apparatus executing arithmetic by a neural network in which multiple processing layers are hierarchically connected is provided. The arithmetic processing apparatus includes multiple arithmetic blocks corresponding to one of the multiple processing layers. Each of the arithmetic blocks includes a convolution arithmetic portion, an activation portion, a pooling portion, and a normalization portion. The convolution arithmetic portion executes convolution arithmetic processing. The normalization portion executes normalization to a processing result data generated by the pooling portion. The normalization portion includes a first output portion, a second output portion, and a normalization execution portion. The first output portion outputs a first data. The second output portion outputs an addition data as a second data. The normalization execution portion executes normalization to the first data based on the second data.

Abstract: An arithmetic processing apparatus performs arithmetic by a neural network in which multiple processing layers are hierarchically connected. The arithmetic processing apparatus corresponding to one of the multiple processing layers includes a convolution arithmetic portion and a pooling processing portion. The convolution arithmetic portion receives an input data from another of the plurality of processing layers, performs convolution arithmetic to the input data, and in each arithmetic cycle, outputs a part of all convolution arithmetic result data required for single pooling processing. The pooling processing portion performs the single pooling processing to the all convolution arithmetic result data before executing activation processing.