Information processing apparatus, information processing method, and storage medium
AssigneeCanon Kabushiki Kaisha
InventorKoichi Tanji
Life of this patent
Estimating the term…
Abstract
An information processing apparatus configured to optimize an architecture of a network model includes at least one memory storing instructions, and at least one processor that, upon execution of the instructions, is configured to perform a search for the architecture based on data for learning, evaluate a topology of the network model that corresponds to the architecture obtained in a process of the search, and perform control to output a change in an evaluation result of the topology as the search progresses.
Claims12
An information processing apparatus configured to optimize an architecture of a network model, the information processing apparatus comprising:
- at least one memory storing instructions;
- and
- at least one processor that, upon execution of the instructions, is configured to:
- perform a search for the architecture based on data for learning, the data for learning including input data and training data, the input data is input to a target network, the training data is used to calculate an error of an output value obtained as a result of a feedforward calculation, and the error is back-propagated in the network;
- evaluate a topological invariant of the network model corresponding to the architecture obtained in a process of the search;
- and
- perform control to output a change in an evaluation result of the topological invariant as the search progresses,
- wherein the search includes searching for an optimum architecture expression that represents a search target network using connection states between a plurality of contact points and calculations between the plurality of contact points.
An information processing method for optimizing an architecture of a network model, the information processing method comprising:
- performing a search for the architecture based on data for learning, the data for learning including input data and training data, the input data is input to a target network, the training data is used to calculate an error of an output value obtained as a result of a feedforward calculation, and the error is back-propagated in the network;
- evaluating a topological invariant of the network model corresponding to the architecture obtained in a process of the search;
- and
- performing control to output a change in an evaluation result of the topological invariant as the search progresses,
- wherein the search includes searching for an optimum architecture expression is such that represents a search target network using by connection states between a plurality of contact points and calculations between the plurality of contact points.
A non-transitory computer-readable storage medium storing computer-executable instructions that configure one or more processors to perform an information processing method for optimizing an architecture of a network model, the information processing method comprising:
- performing a search for the architecture based on data for learning, the data for learning including input data and training data, the input data is input to a target network, the training data is used to calculate an error of an output value obtained as a result of a feedforward calculation, and the error is back-propagated in the network;
- evaluating a topological invariant of the network model corresponding to the architecture obtained in a process of the search;
- and
- performing control to output a change in an evaluation result of the topological invariant as the search progresses,
- wherein the search includes searching for an optimum architecture expression that represents a search target network using by connection states between a plurality of contact points and calculations between the plurality of contact points.
Dependencies are inferred from the claim text.